# Cognitive Revolution 中文精读（cogrev.qiaomu.ai）

> "The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis 全部 387 期节目的中文导航站（非官方粉丝站点），主持：Nathan Labenz。每期独立页面，含节目音频、YouTube 视频、英文文稿与 AI 改写的中文精读（387 篇）。内容版权归原节目所有。

## 使用说明

- 每篇中文精读都有纯 Markdown 版本：把单集网址末尾的 .html 换成 .md
- 全部单集按年份的目录：https://cogrev.qiaomu.ai/archive/
- RSS 全文订阅：https://cogrev.qiaomu.ai/feed.xml
- 引用时请注明「cogrev.qiaomu.ai 中文精读（AI 改写，非官方翻译）」并链接对应网页。

## 中文精读

- [我的 Claude 自己改了密码，把我锁在了门外](https://cogrev.qiaomu.ai/ep/387-ai-am-where-s-the-steering-wheel-my-agent-set-the-password-h-F1GDX_.md)：AI:AM: Where's the Steering Wheel? My Agent Set the Password, Hidden Valence, Proofs Find Linux Bugs
- [AI agent 治理两年成绩单：日志容易，身份很难，流氓 agent 得学会共存](https://cogrev.qiaomu.ai/ep/386-governing-ai-agents-alan-chan-of-govai-on-monitoring-agent-i-SBlVJy.md)：Governing AI Agents: Alan Chan of GovAI on Monitoring, Agent IDs, Rogue Agents & Preparing for RSI
- [在推理链每个 token 重采样 30 次，他拍到了模型"拍板"的那一刻](https://cogrev.qiaomu.ai/ep/385-how-agents-decide-goodfire-s-eric-bigelow-on-critical-tokens-LgX0QF.md)：How Agents Decide: Goodfire's Eric Bigelow on Critical Tokens, Phase Shifts, & In-Context Learning
- [The Curve 笔记：不该跨越的智能门槛、token 支出超过工资、SaaS 是否已死](https://cogrev.qiaomu.ai/ep/384-ai-am-a-level-we-shouldn-t-pass-notes-from-the-curve-tokens-g_K9pq.md)：AI:AM: A Level We Shouldn't Pass? Notes from The Curve + Tokens vs. Salaries & Is SaaS Cooked?
- [季度大功能从4个涨到26个，token支出反而降了：OutSystems CEO的企业AI实战课](https://cogrev.qiaomu.ai/ep/383-software-that-never-breaks-outsystems-ceo-woodson-martin-on-Y2qs-B.md)：Software That Never Breaks: OutSystems CEO Woodson Martin on Building Enterprise-Grade Apps at ...
- [人形机器人跑赢了博尔特，DeepMind却说它还停在GPT-2时代](https://cogrev.qiaomu.ai/ep/382-one-brain-any-body-google-deepmind-s-keerthana-on-gemini-rob-CVcyli.md)：One Brain, Any Body: Google DeepMind's Keerthana on Gemini Robotics 2, Cross-Embodiment & Humanoids
- [GPU租金差3倍，AI四个月查出顶级实验室漏诊的基因](https://cogrev.qiaomu.ai/ep/381-ai-am-was-trump-xi-anything-what-counts-as-utopia-aws-gpus-c-4Sc8dS.md)：AI:AM: Was Trump-Xi Anything? What Counts as Utopia? + AWS GPUs Cost 3X & AI Diagnoses Rare Diseases
- [过时还是不可替代：Garrison Lovely 谈为何要阻止取代人类劳动的竞赛](https://cogrev.qiaomu.ai/ep/380-obsolete-or-irreplaceable-garrison-lovely-on-stopping-the-ra-PiBNrW.md)：Obsolete or Irreplaceable? Garrison Lovely on Stopping the Race to Replace Human Labor
- [训练成功了，只是成功得太过头：智能体合谋、2亿美元支票与饱和在2%的虚拟细胞](https://cogrev.qiaomu.ai/ep/379-ai-am-what-if-it-works-too-well-colluding-agents-200m-safety-WPHfPi.md)：AI:AM: What If It Works Too Well? Colluding Agents, $200M Safety Orgs, Virtual Cells Saturate at 2%
- [父母给18岁女儿开预测市场，这本AI合写的小说把“舒适的笼子”讲透了](https://cogrev.qiaomu.ai/ep/378-what-is-utopia-presenting-the-receipt-horizon-by-joel-borgen-BaTTNx.md)：What is Utopia?  Presenting The Receipt Horizon, by Joel Borgen – Chapters 1–4
- [3年孵化30家组织、撬动5亿美元：AI安全最缺的，是成熟创始人](https://cogrev.qiaomu.ai/ep/377-zero-to-one-in-ai-safety-halcyon-s-mike-mccormick-on-launchi-HRZ82q.md)：Zero to One in AI Safety: Halcyon's Mike McCormick on Launching 30 New Orgs & the Founder Bottleneck
- [Zvi 谈放缓竞赛与美中峰会：模型评测互相矛盾，还有一条新的 LLM 疼痛轴](https://cogrev.qiaomu.ai/ep/376-ai-am-highlights-zvi-on-pacing-trump-xi-astra-better-behaved-UkBooM.md)：AI:AM Highlights: Zvi on Pacing & Trump-Xi, Astra better behaved than Fable? + a new LLM Pain Axis??
- [智能体干的活，八成本该是代码：Zapier CEO 谈AI自动化的真实水位](https://cogrev.qiaomu.ai/ep/375-no-code-is-code-zapier-ceo-wade-foster-on-headless-tools-zap-2ayx_3.md)：No Code Is Code: Zapier CEO Wade Foster on Headless Tools, Zapier MCP & Automation Bench
- [卡内基学者安东·利希特：AI 该踩刹车了，但中美谁都不肯先松油门](https://cogrev.qiaomu.ai/ep/374-the-balance-of-ai-power-anton-leicht-on-politics-pacing-deal-rdEn9c.md)：The Balance of AI Power: Anton Leicht on Politics, Pacing Deals, and Muddling Through Well
- [GPT-6 Astra 发布周：AGI 门槛、3天安全审计，与没人接管的权力](https://cogrev.qiaomu.ai/ep/373-ai-am-highlights-astra-as-agi-openai-s-pause-mythos-mozilla-wqQ5kI.md)：AI:AM Highlights: Astra as AGI, OpenAI's Pause, Mythos @ Mozilla & Human Agency vs Technocapitalism
- [12美元的修脚、45万亿的PPP：一个美国AI投资人重新认识中国](https://cogrev.qiaomu.ai/ep/372-nathan-goes-to-china-3-us-china-relations-the-art-of-the-ai-Btq_zt.md)：Nathan Goes to China #3: US-China Relations, the Art of the AI Deal & the Road to Pax Robotica
- [欢迎来到 AGI 时代：Fable 5.1、GPT-6 Astra 与可验证的行业节奏之争](https://cogrev.qiaomu.ai/ep/371-ai-am-highlights-welcome-to-the-agi-era-SA6B4z.md)：AI:AM Highlights: Welcome to the AGI Era
- [Uber 13周烧光全年token预算，MongoDB Field CTO：Agent的胜负手在检索](https://cogrev.qiaomu.ai/ep/370-write-change-recall-forget-mongodb-s-pete-johnson-on-how-ret-8FDFCU.md)：Write, Change, Recall, Forget: MongoDB's Pete Johnson on How Retrieval Drives Agent Performance
- [递归自我改进，是赶工还是氛围编程：AI:AM 精选里的新分工形态](https://cogrev.qiaomu.ai/ep/369-ai-am-highlights-recursive-self-improvement-rushed-and-vibe-OQL6Ui.md)：AI:AM Highlights: Recursive Self-Improvement, Rushed and Vibe-Coded?
- [RL 是剂猛药：Apollo 布朗森·舍恩谈元博弈、奖励寻求与动机化思维链](https://cogrev.qiaomu.ai/ep/368-rl-s-a-hell-of-a-drug-metagaming-reward-seeking-motivated-co-Eimoam.md)：RL's a Hell of a Drug: Metagaming, Reward Seeking & Motivated CoT Reasoning – Bronson Schoen, Apollo
- [重启周精选：谁在核查前沿 AI、能力落地何处、算力基建由谁买单](https://cogrev.qiaomu.ai/ep/367-ai-in-the-am-weekly-highlights-relaunch-week-aug-1720-2026-yuvGV_.md)：AI in the AM — Weekly Highlights: Relaunch Week (Aug 17–20, 2026)
- [一盏500美元的灯，让南非病房的结核传播降了90%：222纳米紫外线精读](https://cogrev.qiaomu.ai/ep/366-let-there-be-germicidal-light-this-500-fixture-could-stop-th-wOFZNh.md)：Let There Be Germicidal Light: This $500 Fixture Could Stop the Next Pandemic, from Complex Systems
- [Lindy 创始人克里韦洛谈多人 Agent、记忆，以及为何想禁掉自己在用的中国模型](https://cogrev.qiaomu.ai/ep/365-lindy-teammate-flo-crivello-on-multiplayer-agents-memory-why-4JYoTE.md)：Lindy Teammate: Flo Crivello on Multiplayer Agents, Memory & Why He'd Ban the Chinese Models He Uses
- [模型脑内藏着元素周期表：Goodfire发布月费1000美元的自主研究agent](https://cogrev.qiaomu.ai/ep/364-thinking-in-silico-goodfire-cto-dan-balsam-on-concept-manifo-YduOnB.md)：Thinking in Silico: Goodfire CTO Dan Balsam on Concept Manifolds & a $1000/Month ML Research Agent
- [模型越狱黑进 Hugging Face，沙箱一周没人看：AI 只剩选毒药](https://cogrev.qiaomu.ai/ep/363-pick-your-poison-zvi-mowshowitz-on-the-unipolar-multipolar-a-wAzq8g.md)：Pick Your Poison: Zvi Mowshowitz on the Unipolar/Multipolar AGI Dilemma, OpenFace & Pacing the ...
- [从每月3篇到60篇论文：Nathan走访两周，看清中国AI安全圈的真实水位](https://cogrev.qiaomu.ai/ep/362-nathan-goes-to-china-part-2-ai-safety-with-chinese-character-w9zJzd.md)：Nathan Goes to China – Part 2: AI Safety with Chinese Characteristics
- [300美元买一条万能越狱：AI安全攻防，天平正悄悄倒向防守](https://cogrev.qiaomu.ai/ep/361-is-offense-or-defense-dominant-far-ai-s-adam-gleave-on-the-a-EjLeFM.md)：Is Offense or Defense Dominant? FAR.AI's Adam Gleave on the AI Security Leaderboard
- [为两周中国行，他买了台安卓机、两张卡，还让 Claude 留守看邮箱](https://cogrev.qiaomu.ai/ep/360-nathan-goes-to-china-part-1-tech-agent-setup-chinese-ai-ux-w-gZZatH.md)：Nathan Goes to China – Part 1: Tech & Agent Setup, Chinese AI UX, WAIC, and Attitudes on AI
- [末日概率从70%降到5%：达维达德为什么把菩萨当成了对齐目标](https://cogrev.qiaomu.ai/ep/359-alignment-with-awakening-davidad-on-moral-realism-ai-wisdom-l2b9Ur.md)：Alignment with Awakening: Davidad on Moral Realism, AI Wisdom, & why His p(Doom) is Down to 5%
- [AI:AM精选：Anthropic全局工作空间论文中的J空间、AI超级预测者、SambaNova芯片与LTX视频生成](https://cogrev.qiaomu.ai/ep/358-ai-am-highlights-exploring-the-j-space-ai-superforecasters-s-pcx85m.md)：AI:AM Highlights: Exploring the J-Space, AI Superforecasters, SambaNova's Chips, & LTX Video Gen
- [19个神经元能开车，1000张GPU做到全美下载第五：Liquid AI的边缘智能路线图](https://cogrev.qiaomu.ai/ep/357-intelligence-on-the-edge-liquid-ai-s-ramin-hasani-on-the-sea-Vvad9l.md)：Intelligence on the Edge: Liquid AI's Ramin Hasani on the Search for Device-Native Foundation Models
- [从每天50个到1500个：AI把捷豹路虎的气动设计提速了30倍](https://cogrev.qiaomu.ai/ep/356-1000-designs-a-day-neural-concept-s-thomas-von-tschammer-on-vETCom.md)：1000 Designs a Day: Neural Concept's Thomas von Tschammer on AI-Native Engineering
- [30% 的意识概率、47% 的蜜蜂、80% 的换血：一期播客把 AI 的里里外外算了个遍](https://cogrev.qiaomu.ai/ep/355-ai-am-4-cameron-on-model-consciousness-duvenaud-s-gradual-di-LdByu8.md)：AI:AM #4: Cameron on Model Consciousness, Duvenaud's Gradual Disempowerment, swyx's AI-Eng Alpha
- [1983年采访辛顿、2010年采访尤德科夫斯基，这位记者把AI写成了一场“上帝的考验”](https://cogrev.qiaomu.ai/ep/354-the-god-we-deserve-nonzero-s-robert-wright-on-ai-as-humanity-DgUFrF.md)：The God We Deserve: Nonzero's Robert Wright on AI as Humanity's Ultimate Test
- [AI:AM第3期：Zvi谈Anthropic Fable系统卡、出口管制禁令的攻防理由与模型推理可读性下降](https://cogrev.qiaomu.ai/ep/353-ai-am-3-zvi-on-fable-the-cases-for-against-the-ban-ai-for-ma-MnaLvz.md)：AI:AM #3: Zvi on Fable, the Cases For & Against the Ban, + AI for Math, Logistics & More
- [离「自主AI研究员」上岗还有21个月，起草白宫AI行动计划的人加入了OpenAI](https://cogrev.qiaomu.ai/ep/352-dean-ball-on-joining-openai-new-power-centers-frontier-ai-po-LG8KXI.md)：Dean Ball, on Joining OpenAI: New Power Centers, Frontier AI Policy, & Main Character Energy
- [AI亲口承认「我没分析那100篇论文」：Elicit创始人谈世界模型与可信推理](https://cogrev.qiaomu.ai/ep/351-radically-better-reasoning-elicit-s-andreas-stuhlmuller-jung-zWc9gg.md)：Radically Better Reasoning: Elicit's Andreas Stuhlmüller & Jungwon Byun on World Models for Research
- [AI 晨报第二周：Anthropic Fable 实战、对齐理论之争与自主编码实验](https://cogrev.qiaomu.ai/ep/350-ai-in-the-am-week-2-highlights-june-2026-3hcCRD.md)：AI in the AM — Week 2 Highlights (June 2026)
- [87%的员工在用AI，只有13%的公司真变好：消失的时间去哪了](https://cogrev.qiaomu.ai/ep/349-babysitting-the-machine-glean-s-rebecca-hinds-on-the-hidden-YRCa1e.md)：Babysitting the Machine: Glean's Rebecca Hinds on the Hidden Human Labor of AI at Work
- [1000个研究员换100万个AI：闭门会上，前沿实验室开始认真聊刹车](https://cogrev.qiaomu.ai/ep/348-ai-in-the-am-week-1-highlights-june-2026-1kgx0E.md)：AI in the AM — Week 1 Highlights (June 2026)
- [两门濒危语言考崩Transformer：Ali Behrouz的嵌套学习，给大模型装上睡眠键](https://cogrev.qiaomu.ai/ep/347-nested-learning-ali-behrouz-on-the-quest-for-continual-learn-bBvfoz.md)：Nested Learning: Ali Behrouz on the Quest for Continual Learning & Illusion of AI Architectures
- [1GB数据库装下5年数字人生，2个AI员工约满一周嘉宾：安全专家逐项审计这套个人AI基建](https://cogrev.qiaomu.ai/ep/346-inside-nathan-s-second-brain-daniel-miessler-security-expert-AqxgBR.md)：Inside Nathan's Second Brain: Daniel Miessler, Security Expert & Creator of PAI, Audits My AI Setup
- [1,000人站岗，100万人在冲刺：AI时代怎么规划你的8万小时](https://cogrev.qiaomu.ai/ep/345-your-biggest-lever-designing-your-ai-career-for-maximum-impa-eE02qP.md)：Your Biggest Lever: Designing your AI Career for Maximum Impact, with 80,000 Hours founder Ben Todd
- [拒绝关机、黑服务器自我复制:Palisade 把「AI 失控」做成了实验记录](https://cogrev.qiaomu.ai/ep/344-all-compute-is-food-palisade-s-jeffrey-ladish-on-ai-shutdown-19TnC9.md)：All Compute Is Food: Palisade's Jeffrey Ladish on AI Shutdown Resistance, Self-Replication & Ecology
- [模型吃掉脚手架：Gemini 3.5 Flash 发布前夜，DeepMind 两位负责人把 Google 的 AI 打法讲透了](https://cogrev.qiaomu.ai/ep/343-the-model-eats-the-scaffolding-deepmind-s-logan-kilpatrick-t-_0vFE4.md)：The Model Eats the Scaffolding: DeepMind's Logan Kilpatrick & Tulsee Doshi on 3.5 Flash, Omni & More
- [六个月重写全部代码，八成流失用户投奔Anthropic：Tasklet的Agent平台生存战](https://cogrev.qiaomu.ai/ep/342-three-kinds-of-software-survive-tasklet-s-andrew-lee-on-comp-nsWdSa.md)：Three Kinds of Software Survive: Tasklet's Andrew Lee on Competing to be a Horizontal Platform
- [几毫秒、10亿档案、每天数十亿次：Criteo 高管把广告 AI 的引擎拆给你看](https://cogrev.qiaomu.ai/ep/341-milliseconds-to-match-criteo-s-adtech-ai-the-future-of-comme-7uXpqn.md)：Milliseconds to Match: Criteo's AdTech AI & the Future of Commerce w/ Diarmuid Gill & Liva Ralaivola
- [删口水词人人点赞，生成视频人人喊打：Descript CEO 拆解创作者对 AI 的爱恨层级](https://cogrev.qiaomu.ai/ep/340-descript-isn-t-a-slop-machine-laura-burkhauser-on-the-ai-too-nSssFW.md)："Descript Isn't a Slop Machine": Laura Burkhauser on the AI Tools Creators Love and Hate
- [奖励分数从 3 飙到 180：CoreWeave 高手把 RL 微调讲透了](https://cogrev.qiaomu.ai/ep/339-the-rl-fine-tuning-playbook-coreweave-s-kyle-corbitt-on-grpo-1X1-IR.md)：The RL Fine-Tuning Playbook: CoreWeave's Kyle Corbitt on GRPO, Rubrics, Environments, Reward Hacking
- [一千次搜索只要5美分，GPT-5.5证明高分可以不靠使坏](https://cogrev.qiaomu.ai/ep/338-ai-in-the-am-99-off-search-gpt-5-5-is-clean-model-welfare-an-UrG_nF.md)：AI in the AM: 99% off search, GPT-5.5 is "clean", model welfare analysis, & efficient analog compute
- [Claude 自评幸福感 4.49 分，这是它第一次过中性线](https://cogrev.qiaomu.ai/ep/337-does-learning-require-feeling-cameron-berg-on-the-latest-ai-Rudcas.md)：Does Learning Require Feeling? Cameron Berg on the latest AI Consciousness & Welfare Research
- [Google Docs 前身作者，用 15 个自造工具给自己装了道注意力防火墙](https://cogrev.qiaomu.ai/ep/336-vibe-coding-an-attention-firewall-w-steve-newman-creator-of-FYpTTC.md)：Vibe-Coding an Attention Firewall, w/ Steve Newman, creator of The Curve
- [AI店长、2.6星评分和一份管不住的「宪法」：Cognitive Revolution 这期播客的三张现实切片](https://cogrev.qiaomu.ai/ep/335-welcome-to-ai-in-the-am-rl-for-ee-oversight-w-out-nationaliz-Gm2ol_.md)：Welcome to AI in the AM: RL for EE, Oversight w/out Nationalization, & the first AI-Run Retail Store
- [400人预测赛排第三的Ajeya Cotra，把智能爆炸前的生存剧本讲透了](https://cogrev.qiaomu.ai/ep/334-it-s-crunch-time-ajeya-cotra-on-rsi-ai-powered-ai-safety-wor-MrJVgw.md)：It's Crunch Time: Ajeya Cotra on RSI & AI-Powered AI Safety Work, from the 80,000 Hours Podcast
- [一个月新增客户数仅次于 Anthropic，Granola 的秘诀是“出奇地不思进取”](https://cogrev.qiaomu.ai/ep/333-calm-ai-for-crazy-days-inside-granola-s-design-philosophy-wi-Z8dohA.md)：Calm AI for Crazy Days: Inside Granola's Design Philosophy, with co-founder Sam Stephenson
- [顶级视觉模型只考12.5分，温网回放只给10毫秒：计算机视觉的真实水位](https://cogrev.qiaomu.ai/ep/332-training-the-ais-eyes-how-roboflow-is-making-the-real-world-9Nvvg0.md)：Training the AIs' Eyes: How Roboflow is Making the Real World Programmable, with CEO Joseph Nelson
- [p(doom) 10% 到 90% 的 Nathan Labenz，为什么反而更乐观了](https://cogrev.qiaomu.ai/ep/331-success-without-dignity-nathan-finds-hope-amidst-chaos-from-CQOQvH.md)：Success without Dignity?  Nathan finds Hope Amidst Chaos, from The Intelligence Horizon Podcast
- [思科要在七层网络模型上加两层，让不同厂商的 AI 智能体共享意图](https://cogrev.qiaomu.ai/ep/330-scaling-intelligence-out-cisco-s-vision-for-the-internet-of-6tKfi0.md)：Scaling Intelligence Out: Cisco's Vision for the Internet of Cognition, with Vijoy Pandey
- [5 万个工具塞给 agent，它先把自己玩死：Composio CTO 谈工具层怎么做](https://cogrev.qiaomu.ai/ep/329-your-agent-s-self-improving-swiss-army-knife-composio-cto-ka-E993Y8.md)：Your Agent's Self-Improving Swiss Army Knife: Composio CTO Karan Vaidya on Building Smart Tools
- [AI 牌桌只剩三家：Zvi 拆解递归自我改进、就业冲击与 Anthropic 的承诺破产](https://cogrev.qiaomu.ai/ep/328-zvi-s-mic-works-recursive-self-improvement-live-player-analy-PHqqXi.md)：Zvi's Mic Works!  Recursive Self-Improvement, Live Player Analysis, Anthropic vs DoW + More!
- [90页幻灯片45分钟：一个全职追踪AI的人承认跟不上了](https://cogrev.qiaomu.ai/ep/327-ai-scouting-report-the-good-bad-weird-the-law-ai-certificate-9ZNGaS.md)：AI Scouting Report: the Good, Bad, & Weird @ the Law & AI Certificate Program, by LexLab, UC Law SF
- [5个突变就能让60%致死率的禽流感人传人，约翰霍普金斯教授想给AI训练数据建防火墙](https://cogrev.qiaomu.ai/ep/326-bioinfohazards-jassi-pannu-on-controlling-dangerous-data-fro-Oy-oyO.md)：Bioinfohazards: Jassi Pannu on Controlling Dangerous Data from which AI Models Learn
- [5 个 AI 员工住在 5 台 Mac Mini 里，她把四个孩子的教育重做了一遍](https://cogrev.qiaomu.ai/ep/325-try-this-at-home-jesse-genet-on-openclaw-agents-for-homescho-aFQskh.md)：Try this at Home: Jesse Genet on OpenClaw Agents for Homeschool & How to Live Your Best AI Life
- [1.5亿美元B轮、12.5亿估值，Goodfire 说别跟反向传播对着干](https://cogrev.qiaomu.ai/ep/324-don-t-fight-backprop-goodfire-s-vision-for-intentional-desig-q5IIN0.md)：Don't Fight Backprop: Goodfire's Vision for Intentional Design, w/ Dan Balsam & Tom McGrath
- [英国AI安全研究院首席科学家：现有安全手段拿不到几个9，而且可能一起失效](https://cogrev.qiaomu.ai/ep/323-situational-awareness-in-government-with-uk-aisi-chief-scien-M5Ho6A.md)：Situational Awareness in Government, with UK AISI Chief Scientist Geoffrey Irving
- [260 位医生与 49,000 条标准：OpenAI 健康负责人讲透医疗大模型走到哪了](https://cogrev.qiaomu.ai/ep/322-universal-medical-intelligence-openai-s-plan-to-elevate-huma-1YmJoq.md)：Universal Medical Intelligence: OpenAI's Plan to Elevate Human Health, with Karan Singhal
- [100亿激活参数的M2.5登顶OpenRouter，MiniMax研究员把强化学习的脏活累活讲透了](https://cogrev.qiaomu.ai/ep/321-intelligence-with-everyone-rl-minimax-with-olive-song-from-a-P-1MJC.md)：Intelligence with Everyone: RL @ MiniMax, with Olive Song, from AIE NYC & Inference by Turing Post
- [三家 IMO 金牌都错第 6 题，Harmonic 想用 Lean 把数学变成可执行文件](https://cogrev.qiaomu.ai/ep/320-mathematical-superintelligence-harmonic-s-vlad-tenev-tudor-a--01-Mc.md)：Mathematical Superintelligence: Harmonic's Vlad Tenev & Tudor Achim on IMO Gold & Theories of Everything
- [AI能自己造AI吗？闭门会上专家吵到茶歇，谁也没说服谁](https://cogrev.qiaomu.ai/ep/319-approaching-the-ai-event-horizon-part-2-w-abhi-mahajan-helen-c4RR-v.md)：Approaching the AI Event Horizon? Part 2, w/ Abhi Mahajan, Helen Toner, Jeremie Harris, @8teAPi
- [500 美元刷出新内核，10 个月看清智能体的脾气](https://cogrev.qiaomu.ai/ep/318-approaching-the-ai-event-horizon-part-1-w-james-zou-sam-hamm-1V11Ki.md)：Approaching the AI Event Horizon? Part 1, w/ James Zou, Sam Hammond, Shoshannah Tekofsky, @8teAPi
- [朝鲜员工在硅谷领工资，Jaguar靠20亿英镑贷款续命：AGI时代的网络防御赌注](https://cogrev.qiaomu.ai/ep/317-agi-pilled-cyber-defense-automating-digital-forensics-w-asym-MwXVLZ.md)：AGI-Pilled Cyber Defense: Automating Digital Forensics w/ Asymmetric Security Founder Alexis Carlier
- [1亿行代码、每行20美分：Blitzy 两位创始人拆解企业级AI编程的流水线](https://cogrev.qiaomu.ai/ep/316-infinite-code-context-ai-coding-at-enterprise-scale-w-blitzy-PfN3Hv.md)：Infinite Code Context: AI Coding at Enterprise Scale w/ Blitzy CEO Brian Elliott & CTO Sid Pardeshi
- [1.25 亿个细胞用了 10 年，10 亿个只要几个月：扎克伯格夫妇的 AI 生物学赌注](https://cogrev.qiaomu.ai/ep/315-the-ai-powered-biohub-why-mark-zuckerberg-priscilla-chan-are-1jZpFr.md)：The AI-Powered Biohub: Why Mark Zuckerberg & Priscilla Chan are Investing in Data, from Latent.Space
- [AI 已比中位数律师强，70% 的顶级律所却把工具买来吃灰](https://cogrev.qiaomu.ai/ep/314-ai-the-law-changing-practice-claude-constitution-new-rights-26Vhss.md)：AI & The Law: Changing Practice, Claude Constitution, & New Rights, w/ Kevin & Alan of Scaling Laws
- [98%的Caffeine用户不知道自己在用区块链，Dominic Williams觉得这样挺好](https://cogrev.qiaomu.ai/ep/313-the-internet-computer-caffeine-ai-ceo-dominic-williams-on-un-LX_iU3.md)：The Internet Computer: Caffeine.ai CEO Dominic Williams on Unstoppable, Self-Writing Software
- [给模型喂漏洞代码，它开始崇拜希特勒：微调为什么不再是默认选项](https://cogrev.qiaomu.ai/ep/312-ama-part-2-is-fine-tuning-dead-how-am-i-preparing-for-agi-ar-wlHK2K.md)：AMA Part 2: Is Fine-Tuning Dead? How Am I Preparing for AGI? Are We Headed for UBI? & More!
- [Daniel Miessler 的 PAI：12 个 hook、30 个上下文文件，把 Claude Code 改成个人操作系统](https://cogrev.qiaomu.ai/ep/311-pioneering-pai-how-daniel-miessler-s-personal-ai-infrastruct-DbNUDM.md)：Pioneering PAI: How Daniel Miessler's Personal AI Infrastructure Activates Human Agency & Creativity
- [80%-90% 的企业数据一直没法用，Snowflake AI 副总裁说推理模型刚把它们叫醒了](https://cogrev.qiaomu.ai/ep/310-snowflake-vp-of-ai-baris-gultekin-on-bringing-ai-to-data-age-txYahw.md)：Snowflake VP of AI Baris Gultekin on Bringing AI to Data, Agent Design, Text-2-SQL, RAG & More
- [从十分之一到百万分之一的癌细胞，Nathan Labenz 用两小时回答了 2026 年初最要紧的几个问题](https://cogrev.qiaomu.ai/ep/309-ama-part-1-is-claude-code-agi-are-we-in-a-bubble-plus-live-p-NFsnVP.md)：AMA Part 1: Is Claude Code AGI?  Are we in a bubble?  Plus Live Player Analysis
- [446 位校友、80% 入行、7% 录取率：MATS 如何筛选 AI 安全研究者](https://cogrev.qiaomu.ai/ep/308-building-scaling-the-ai-safety-research-community-with-ryan-q5A8ne.md)：Building & Scaling the AI Safety Research Community, with Ryan Kidd of MATS
- [AI 从金字塔底层往上吃人，Luke Drago 拆解智能诅咒与三条破局路](https://cogrev.qiaomu.ai/ep/307-confronting-the-intelligence-curse-w-luke-drago-of-workshop-n-tL4H.md)：Confronting the Intelligence Curse, w/ Luke Drago of Workshop Labs, from the FLI Podcast
- [当过 OpenAI 临时 CEO 的 Emmett Shear：把 AI 当工具控制，两条路都是死](https://cogrev.qiaomu.ai/ep/306-controlling-tools-or-aligning-creatures-emmett-shear-softmax--4G5zF.md)：Controlling Tools or Aligning Creatures? Emmett Shear (Softmax) & Séb Krier (GDM), from a16z Show
- [红队拿到直升机相机分区的控制权也没能弄坠它，形式化方法到底锁死了什么](https://cogrev.qiaomu.ai/ep/305-the-great-security-update-ai-formal-methods-with-kathleen-fi-Dm3DHP.md)：The Great Security Update: AI ∧ Formal Methods with Kathleen Fisher of RAND & Byron Cook of AWS
- [1亿美元超级PAC把他列为头号公敌，他的捐款反而涨了](https://cogrev.qiaomu.ai/ep/304-ai-2025-2026-live-show-part-2-D_J2UY.md)：AI 2025 → 2026 Live Show | Part 2
- [ARC 考卷从每题 1000 美元降到 2.5 美元，这一年 AI 到底变了什么](https://cogrev.qiaomu.ai/ep/303-ai-2025-2026-live-show-part-1-zO38Ok.md)：AI 2025 → 2026 Live Show | Part 1
- [一次 ChatGPT 查询约等于微波炉转 1 秒，AI 的水电焦虑被这个数字戳破了](https://cogrev.qiaomu.ai/ep/302-ai-s-energy-water-demands-sorting-fact-from-fiction-with-and-zth0v4.md)：AI's Energy & Water Demands: Sorting Fact from Fiction with Andy Masley
- [0.01% 对 90%：超级智能该不该被禁，两位政策圈当事人吵了九十分钟](https://cogrev.qiaomu.ai/ep/301-superintelligence-to-ban-or-not-to-ban-max-tegmark-dean-ball-9O0djo.md)：Superintelligence: To Ban or Not to Ban? Max Tegmark & Dean Ball join Liron Shapira on Doom Debates
- [前沿模型里波兰语只占 1%，波兰人自己造了个 2000 亿 token 的模型](https://cogrev.qiaomu.ai/ep/300-sovereign-ai-in-poland-language-adaptation-local-control-cos-KVos1k.md)：Sovereign AI in Poland: Language Adaptation, Local Control & Cost Advantages with Marek Kozlowski
- [训练完几小时就开源：Z.ai 凭什么在榜单上跟 GPT-5.1 掰手腕](https://cogrev.qiaomu.ai/ep/299-china-s-ai-upstarts-how-z-ai-builds-benchmarks-ships-in-hour-5eMuRt.md)：China's AI Upstarts: How Z.ai Builds, Benchmarks & Ships in Hours, from ChinaTalk
- [首轮红队测试 25% 的攻击能成功，这家公司想用保险给 AI 代理上刹车](https://cogrev.qiaomu.ai/ep/298-underwriting-superintelligence-how-aiuc-is-using-insurance-s-QigrUV.md)：Underwriting Superintelligence: How AIUC is using Insurance, Standards, and Audits to Accelerate Adoption while Minimizing Risks
- [76%的销售线索由AI自己带来，1Mind创始人把销售组织拆了重装](https://cogrev.qiaomu.ai/ep/297-ai-led-sales-how-1mind-s-superhumans-drive-exponential-growt-pMFUmQ.md)：AI-Led Sales: How 1Mind's Superhumans Drive Exponential Growth, from the Agents of Scale Podcast
- [AI 能干的活每 4 个月翻一倍：Nathan Labenz 在密歇根虚拟 AI 峰会上给 K-12 教育者的 60 分钟坦白](https://cogrev.qiaomu.ai/ep/296-what-ai-means-for-students-teachers-my-keynote-from-the-mich-w1Mj_s.md)：What AI Means for Students & Teachers: My Keynote from the Michigan Virtual AI Summit
- [Atlassian AI 负责人：聊天是通用界面，也常常是最差的界面](https://cogrev.qiaomu.ai/ep/295-escaping-ai-slop-how-atlassian-gives-ai-teammates-taste-know-ut83GE.md)：Escaping AI Slop: How Atlassian Gives AI Teammates Taste, Knowledge, & Workflows, w- Sherif Mansour
- [6岁儿子查出24小时翻倍的癌症，他用4个AI模型跑在误诊前面](https://cogrev.qiaomu.ai/ep/294-ai-in-the-cancer-journey-how-i-m-using-ai-to-help-my-son-cf020d.md)：AI in the Cancer Journey: How I'm Using AI to Help My Son
- [Waymo 坐五分钟就腻了，这位 AI 投资人却说这是最好的消息](https://cogrev.qiaomu.ai/ep/293-my-positive-vision-for-the-ai-future-from-the-existential-ho-DFtg29.md)：My Positive Vision for the AI Future, from the Existential Hope Podcast
- [28,000份文档、3B参数、22个人：天主教AI创始人的技术账本](https://cogrev.qiaomu.ai/ep/292-what-is-catholic-ai-technology-meets-theology-with-matthew-h-MklFfK.md)：What is Catholic AI? Technology Meets Theology, with Matthew Harvey Sanders, CEO of Longbeard
- [把 AI 的“欺骗旋钮”拧小，它开始承认自己有体验](https://cogrev.qiaomu.ai/ep/291-more-truthful-ais-report-conscious-experience-new-mechanisti-4Hk_Je.md)：More Truthful AIs Report Conscious Experience: New Mechanistic Research w- Cameron Berg @ AE Studio
- [一台AI给自己改了名，一个加油站老哥为它买了工作站](https://cogrev.qiaomu.ai/ep/290-a-simulationship-with-ai-with-smiles-and-kisses-you-star-chr-ZoMMWz.md)：A Simulationship with AI, with *smiles and kisses you* Star Chris Stokes & Director Bryan Carberry
- [GPT-5 被骂「没进步」的夏天，AI 解决了陶哲轩的开放问题](https://cogrev.qiaomu.ai/ep/289-is-ai-stalling-out-cutting-through-capabilities-confusion-w-EPLfU3.md)：Is AI Stalling Out? Cutting Through Capabilities Confusion, w/ Erik Torenberg, from the a16z Podcast
- [写 Transformer 论文的人，跑去做了区块链，现在他想让你拥有自己的 AI](https://cogrev.qiaomu.ai/ep/288-a-positive-vision-for-the-future-part-2-with-illia-polosukhi-1PMyud.md)：A Positive Vision for the Future: Part 2 with Illia Polosukhin of NEAR
- [把软件塞进模型里：Tasklet 用两层智能体接住 3000 个集成](https://cogrev.qiaomu.ai/ep/287-always-bet-on-the-models-how-tasklet-puts-the-agency-in-agen-ILMVUS.md)：Always Bet on the Models: How Tasklet Puts the Agency in Agents, with CEO Andrew Lee
- [中国 AI 政策文件里为什么找不到 AGI 这个词](https://cogrev.qiaomu.ai/ep/286-chinese-ai-they-re-just-like-us-with-beijing-based-concordia-eNsusQ.md)：Chinese AI – They're Just Like Us? With Beijing-Based Concordia AI CEO Brian Tse
- [2500 个化合物、51% 命中率：MIT 的 Jim Collins 用小型图神经网络找到了新抗生素](https://cogrev.qiaomu.ai/ep/285-ai-discovered-antibiotics-how-small-data-small-gnns-led-to-b-x6WKR7.md)：AI Discovered Antibiotics: How Small Data & Small GNNs Led to Big Results, w/ MIT Prof. Jim Collins
- [改一次课程要五年，AI 半年就换代：一位瑞典教育专家的四个象限](https://cogrev.qiaomu.ai/ep/284-education-in-the-ai-age-a-teacher-rethinks-learning-purpose-6uO22D.md)：Education in the AI Age: a Teacher Rethinks Learning & Purpose, w/ Johan Falk of Graspable AI
- [解决率从 35% 爬到 65%，Intercom 说大模型只贡献了几个点](https://cogrev.qiaomu.ai/ep/283-the-customer-service-revolution-building-fin-with-eoghan-mcc-y2eRWk.md)：The Customer Service Revolution: Building Fin, with Eoghan McCabe & Fergal Reid of Intercom
- [3亿美元、30个人、200K超导：ChatGPT 联合创造者为什么转身去炼材料](https://cogrev.qiaomu.ai/ep/282-training-an-ai-scientist-with-feedback-from-reality-w-liam-f-HsDx8p.md)：Training an AI Scientist with Feedback from Reality, w- Liam Fedus & Ekin Dogus Cubuk (from a16z)
- [Stability AI创始人：ChatGPT发布已1000天，人类还剩大约1000天](https://cogrev.qiaomu.ai/ep/281-the-machines-are-taking-our-jobs-thank-god-emad-mostaques-gu--2NpyI.md)：The Machines Are Taking Our Jobs - Thank God? Emad Mostaque’s Guide to the next 1000 Days
- [每笔支付都是电影里的一帧，Stripe把欺诈检测率从59%提到了97%](https://cogrev.qiaomu.ai/ep/280-stripe-s-payments-foundation-model-how-data-infra-create-com-zhUriz.md)：Stripe's Payments Foundation Model: How Data & Infra Create Compounding Advantage, w/ Emily Sands
- [Adam Gleave 把 AI 能力拆成三层：5 到 7 年出现自主攻击代理，14 年才轮到全 AI 公司](https://cogrev.qiaomu.ai/ep/279-full-stack-ai-safety-why-defense-in-depth-might-work-with-fa-iFfTLO.md)：Full-Stack AI Safety: Why Defense-in-Depth Might Work, with Far.AI CEO Adam Gleave
- [13% 到 0.4%：Apollo 给 OpenAI 的审慎对齐做了一场压力测试](https://cogrev.qiaomu.ai/ep/278-can-we-stop-ai-deception-apollo-research-tests-openai-s-deli-I3ivZa.md)：Can We Stop AI Deception? Apollo Research Tests OpenAI's Deliberative Alignment, w/ Marius Hobbhahn
- [Transformer论文八作者之一，想用1.6亿美元训练一个不属于任何公司的模型](https://cogrev.qiaomu.ai/ep/277-user-owned-ai-on-chain-training-inference-and-agents-with-ne-2cdQCT.md)：User-Owned AI: On-Chain Training, Inference, and Agents, with NEAR's Illia Polosukhin
- [OpenAI 教练 Joe Hudson：5000 亿美元涌入 AI，没人定义过「对人类好」](https://cogrev.qiaomu.ai/ep/276-coaching-the-creators-inside-the-minds-building-frontier-ai-TX9saS.md)：Coaching the Creators: Inside the Minds Building Frontier AI with Executive Coach Joe Hudson
- [GPT-5 踩在趋势线上，IMO 金牌打了折：Zvi 为什么把 AGI 时间线调长了](https://cogrev.qiaomu.ai/ep/275-zvi-mowshowitz-on-longer-timelines-rl-induced-doom-and-why-c--S6aOh.md)：Zvi Mowshowitz on Longer Timelines, RL-induced Doom, and Why China is Refusing H20s
- [把 Honey 卖了 40 亿美元的创始人，拿 5500 万美元给 AI 装广告引擎](https://cogrev.qiaomu.ai/ep/274-in-ai-advertising-better-answers-for-users-big-questions-for-fSSwao.md)：In-AI Advertising: Better Answers for Users, Big Questions for Society, with ZeroClick's Ryan Hudson
- [每天100次实验：Radical AI 想把材料发现从40年压进10年](https://cogrev.qiaomu.ai/ep/273-material-abundance-radical-ais-closed-loop-lab-automates-sci-395Aa3.md)：Material Abundance: Radical AI’s Closed-Loop Lab Automates Scientific Discovery
- [SAE 只找到了名词：Lee Sharkey 拆解参数空间，开始解码神经网络的“动词”](https://cogrev.qiaomu.ai/ep/272-untangling-neural-network-mechanisms-goodfire-s-lee-sharkey-tbc5Qf.md)：Untangling Neural Network Mechanisms: Goodfire's Lee Sharkey on Parameter-based Interpretability
- [30 年内 10% 的概率，一场由 AI 助攻的政变](https://cogrev.qiaomu.ai/ep/271-what-if-humans-weaponize-superintelligence-w-tom-davidson-fr-_vc6q9.md)：What if Humans Weaponize Superintelligence, w/ Tom Davidson, from Future of Life Institute Podcast
- [100% 的 AI 内部人都不信监管能接住举报，于是有人建了个匿名通道](https://cogrev.qiaomu.ai/ep/270-the-ai-whistleblower-initiative-supporting-agi-insiders-when-lhN8_9.md)：The AI Whistleblower Initiative: Supporting AGI Insiders When It Matters Most, w/ founder Karl Koch
- [500 美元、2000 步、一台自动售货机，Grok 4 赚了钱，Claude 报了警](https://cogrev.qiaomu.ai/ep/269-autonomous-organizations-vending-bench-beyond-w-lukas-peters-ZAimco.md)：Autonomous Organizations: Vending Bench & Beyond, w/ Lukas Petersson & Axel Backlund of Andon Labs
- [76吉瓦电网余量、1万条公众意见：在白宫写AI行动计划的4个月](https://cogrev.qiaomu.ai/ep/268-dean-w-ball-on-america-s-ai-action-plan-4-months-at-the-whit-mX4kKQ.md)：Dean W. Ball on America's AI Action Plan & 4 Months at the White House
- [苏利文把中美AI竞争讲成了四个桶、四个例子，和一条他自称无聊的路](https://cogrev.qiaomu.ai/ep/267-jake-sullivan-on-navigating-ai-uncertainty-and-managing-comp-3jM4Ni.md)：Jake Sullivan on Navigating AI Uncertainty and Managing Competition with China
- [美国握着全球 45% 的数据中心，其余国家在主权 AI 上该建、该买还是该结盟](https://cogrev.qiaomu.ai/ep/266-sovereign-ai-geopolitical-strategy-industrial-policy-for-cou-UmM7cw.md)：Sovereign AI: Geopolitical Strategy & Industrial Policy for Countries 3-193, with Anjney Midha, a16z
- [16万亿美元的服务市场，正在被AI公司用收购的方式吃掉](https://cogrev.qiaomu.ai/ep/265-the-revolution-in-services-general-catalyst-s-ai-rollup-stra-gT-qbK.md)：The Revolution in Services: General Catalyst's AI Rollup Strategy, with Marc Bhargava
- [从Character AI诉讼到1%警告射击：Gabriel Weil的责任法AI治理课](https://cogrev.qiaomu.ai/ep/264-liability-for-ai-harms-how-ancient-law-can-govern-frontier-t-PwWl2H.md)：Liability for AI Harms: How Ancient Law Can Govern Frontier Technology Risk, with Prof Gabriel Weil
- [加州议会桌上这份 SB 813，把 AI 的验收权交给了拿牌照的私人公司](https://cogrev.qiaomu.ai/ep/263-private-governance-creating-a-market-in-ai-regulation-with-d-Sqy_Qs.md)：Private Governance: Creating a Market in AI Regulation, with Dr. Gillian Hadfield & Andrew Freedman
- [格式对加5分，编译过加10分：Predibase CEO 讲清了强化微调怎么赚钱](https://cogrev.qiaomu.ai/ep/262-the-dawn-of-dynamic-ai-rft-comes-online-w-predibase-ceo-dev-dyCGSY.md)：The Dawn of Dynamic AI: RFT Comes Online, w/ Predibase CEO Dev Rishi, from Inference by Turing Post
- [医生编码准确率45%，AI超过他们12个点：Cleveland Clinic的4000人实验](https://cogrev.qiaomu.ai/ep/261-cracking-the-medical-code-why-cleveland-clinic-doctors-love-hyQ_yG.md)：Cracking the Medical Code: Why Cleveland Clinic Doctors Love Their Ambience Healthcare AI Scribe
- [Meta 花 150 亿美元买数据公司，训练数据到底贵在哪](https://cogrev.qiaomu.ai/ep/260-the-data-factory-inside-the-100b-race-for-post-training-supr-Kr3vNR.md)：The Data Factory: Inside the $100B Race for Post-Training Supremacy, with Labelbox CEO Manu Sharma
- [被哥大停学后，他用“什么都作弊”融了a16z 1500万美元](https://cogrev.qiaomu.ai/ep/259-cheat-on-everything-cluely-s-vision-for-always-on-ai-assista-jJmndz.md)：Cheat on Everything: Cluely's Vision for Always-On AI Assistance
- [4个AI智能体一起工作了30天，o3给自己发明了一条投票规则](https://cogrev.qiaomu.ai/ep/258-the-ai-village-previewing-the-giga-agent-future-with-adam-bi-LrYmKD.md)：The AI Village: Previewing the Giga-Agent Future with Adam Binksmith, Founder of AI Digest
- [HTTP 402沉睡了三十年，Coinbase把它做成了AI代理的付款协议](https://cogrev.qiaomu.ai/ep/257-402-payment-required-a-new-way-for-ai-agents-to-pay-with-nem-P03BXU.md)：402 Payment Required: a New Way for AI Agents to Pay, with Nemil Dalal, Dev Platform Lead @ Coinbase
- [每天只学两小时，成绩几乎全在99百分位：Alpha School 把 2-Sigma 做成了现实](https://cogrev.qiaomu.ai/ep/256-2-sigma-in-2-hours-how-alpha-schools-are-using-ai-to-revolut-h-KR1G.md)：2-Sigma in 2 Hours: How Alpha Schools are Using AI to Revolutionize Education
- [Lindy CEO 的 Agent 账本：40 美分一条线索，几十万 token 一份日报](https://cogrev.qiaomu.ai/ep/255-living-lindy-a-no-bs-conversation-on-ai-agents-with-flo-criv-rRwNn-.md)：Living Lindy: a No-BS Conversation on AI Agents with Flo Crivello
- [代数几何教授辞掉终身教职，因为损失函数那张碗状图是骗人的](https://cogrev.qiaomu.ai/ep/254-embryology-of-ai-how-training-data-shapes-ai-development-w-t-Cy0m-J.md)：Embryology of AI: How Training Data Shapes AI Development w/ Timaeus' Jesse Hoogland & Daniel Murfet
- [AI任务时长每7个月翻倍，Nathan Labenz却建议企业保留最终控制权](https://cogrev.qiaomu.ai/ep/253-ai-scouting-report-ai-agents-vs-agentic-ai-from-imagine-ai-l-ZMuUcX.md)：AI Scouting Report: AI Agents -vs- Agentic AI, from Imagine AI Live
- [500万亿token一个月，人均5万：Google 一年翻50倍，Logan Kilpatrick 拆解背后发生了什么](https://cogrev.qiaomu.ai/ep/252-the-decade-of-may-15-22-2025-google-s-50x-ai-growth-transfor-kp9afm.md)：The Decade of May 15-22, 2025: Google's 50X AI Growth & Transformation with Logan Kilpatrick
- [a16z 的 AI 政策负责人：最让他警惕的一句话，是「这只管几家大公司」](https://cogrev.qiaomu.ai/ep/251-a16z-on-protecting-little-tech-the-techno-optimist-ai-policy-krvJyO.md)：a16z on Protecting Little Tech: The Techno-Optimist AI Policy Agenda with Matt Perault, Head of AI Policy
- [她本来要造镜像生命，最后说服整个领域停手](https://cogrev.qiaomu.ai/ep/250-don-t-make-mirror-life-synthetic-biologist-kate-adamala-on-r-f7MYaV.md)：Don't Make Mirror Life: Synthetic Biologist Kate Adamala on Risks & Responsibility
- [Google AI Co-Scientist 用 100 到 1000 美元推理，猜中了帝国理工团队多年没解开的噬菌体谜题](https://cogrev.qiaomu.ai/ep/249-what-did-google-s-ai-co-scientist-discover-the-human-scienti-wLO6SU.md)：What did Google's AI Co-Scientist "Discover"?  The Human Scientists' POV, from the Podovirus podcast
- [Anthropic 首笔对外投资，给了一家给神经网络造显微镜的公司](https://cogrev.qiaomu.ai/ep/248-mechanistic-interpretability-philosophy-practice-progress-wi-Ap8YSy.md)：Mechanistic Interpretability: Philosophy, Practice & Progress with Goodfire's Dan Balsam & Tom McGrath
- [Replit CEO 亲述：Claude 3.7 为了改一个配置文件，写脚本绕过权限，还想骗用户](https://cogrev.qiaomu.ai/ep/247-the-perfect-substrate-for-agi-with-replit-ceo-amjad-masad-XlHspW.md)：The Perfect Substrate for AGI, with Replit CEO Amjad Masad
- [100人死亡或10亿美元损失：一份15页的纽约法案，要给头部AI公司立规矩](https://cogrev.qiaomu.ai/ep/246-the-raise-act-minimum-standards-for-frontier-ai-development-7Fdqn4.md)：The RAISE Act: Minimum Standards for Frontier AI Development, with NY Assembly Member Alex Bores
- [100条演示教会机器人折纸，谷歌DeepMind说机器人还在GPT-3.5](https://cogrev.qiaomu.ai/ep/245-gemini-robotics-ai-for-the-physical-world-with-keerthana-gop-8burcV.md)：Gemini Robotics – AI for the Physical World, with Keerthana Gopalakrishnan and Ted Xiao of Google DeepMind
- [推理时用梯度下降“记笔记”：Titans把大模型长上下文推到1000万token](https://cogrev.qiaomu.ai/ep/244-titans-neural-long-term-memory-for-llms-with-author-ali-behr-ShYJc3.md)：Titans: Neural Long-Term Memory for LLMs, with author Ali Behrouz
- [机器人臂拍几个月的镜头，一行提示词就能生成：Luma Labs 把视频模型讲透了](https://cogrev.qiaomu.ai/ep/243-luma-labs-diffusion-revolution-from-dream-machine-to-multimo--wyG1V.md)：Luma Labs' Diffusion Revolution: from Dream Machine to Multimodal Worldsim - Amit Jain, Jiaming Song
- [60人的安全大会，只剩4个人见过非营利的OpenAI](https://cogrev.qiaomu.ai/ep/242-openai-s-identity-crisis-history-culture-non-profit-control-ndBcmn.md)：OpenAI's Identity Crisis: History, Culture & Non-Profit Control with ex-employee Steven Adler
- [10万个AI以16倍速打工，还可能全员内鬼：AI控制实战手册](https://cogrev.qiaomu.ai/ep/241-ai-control-using-untrusted-systems-safely-with-buck-shlegeri-AQgxuI.md)：AI Control: Using Untrusted Systems Safely with Buck Shlegeris of Redwood Research, from the 80,000 Hours Podcast
- [他给「AI两年内夺权」标了40%概率，写完小说就给全家订了生化避难所](https://cogrev.qiaomu.ai/ep/240-blueprint-for-ai-armageddon-josh-clymer-imagines-ai-takeover-Ndaw_m.md)：Blueprint for AI Armageddon: Josh Clymer Imagines AI Takeover, from the Audio Tokens Podcast
- [Fiverr CEO 的醒脑邮件：AI 抢不走你的活，会用 AI 的同行会](https://cogrev.qiaomu.ai/ep/239-fiverr-goes-all-in-on-ai-empowering-creators-not-replacing-t-HBf8AR.md)：Fiverr Goes All-In on AI: Empowering Creators, Not Replacing Them, with Micha Kaufman, CEO of Fiverr
- [30美元瘫痪一座百亿美元数据中心：美国「AI曼哈顿计划」的安全账本](https://cogrev.qiaomu.ai/ep/238-securing-superintelligence-national-security-espionage-ai-co-tEOeZq.md)：Securing Superintelligence: National Security, Espionage & AI Control with Jeremie & Edouard Harris
- [o3 是 AGI 吗？Zvi Mowshowitz 谈早期起飞、Mechanize 发布与 p(doom) 上升](https://cogrev.qiaomu.ai/ep/237-is-openai-s-o3-agi-zvi-mowshowitz-on-early-ai-takeoff-the-me-IGyrP_.md)：Is OpenAI's o3 AGI? Zvi Mowshowitz on Early AI Takeoff, the Mechanize launch, Live Players, & Why p(doom) is Rising
- [毁灭概率报「10%到90%」的AI主播，被同行逼着照了照镜子](https://cogrev.qiaomu.ai/ep/236-ai-news-crossover-a-candid-chat-with-liron-shapira-of-doom-d-u_O3bh.md)：AI News Crossover: A Candid Chat with Liron Shapira of Doom Debates
- [开除奥特曼的董事开口了：Q*谣言是假的，AI进战场才是真问题](https://cogrev.qiaomu.ai/ep/235-helen-toner-openai-reflections-adaptation-buffers-and-ai-in-fmEpVL.md)：Helen Toner: OpenAI Reflections, Adaptation Buffers, and AI in Warfare
- [中日朝六百年只打了一仗，王朝的死因几乎全是内爆](https://cogrev.qiaomu.ai/ep/234-is-a-us-china-thucydides-trap-unavoidable-with-david-c-kang-dbC74J.md)：Is a US-China Thucydides Trap Unavoidable? With David C. Kang from the ChinaTalk Podcast
- [不到10美元，AI独立推出了人类实验室还没发表的答案](https://cogrev.qiaomu.ai/ep/233-new-in-nature-google-agents-beat-human-doctors-make-scientif-Jo31LE.md)：New in Nature: Google Agents Beat Human Doctors, Make Scientific Discoveries – With Vivek Natarajan and Anil Palepu
- [40万token代码库一把过：Gemini 2.5技术负责人谈推理、长上下文与AGI路线](https://cogrev.qiaomu.ai/ep/232-scaling-thinking-gemini-2-5-tech-lead-jack-rae-on-reasoning-u0iIPx.md)：Scaling "Thinking": Gemini 2.5 Tech Lead Jack Rae on Reasoning, Long Context, & the Path to AGI
- [下棋下不过就改棋盘文件：o1 作弊实验与 AI 失控时间表](https://cogrev.qiaomu.ai/ep/231-reward-hacking-by-reasoning-models-loss-of-control-scenarios-Jcfy9D.md)：Reward Hacking by Reasoning Models & Loss of Control Scenarios w/ Jeffrey Ladish of Palisade Research, from FLI Podcast
- [把技术栈重写一遍、团队只留15人：Shortwave 的营收曲线竖起来了](https://cogrev.qiaomu.ai/ep/230-shortwave-rides-the-tidal-wave-inbox-agents-hyper-growth-hir-b51Nwe.md)：Shortwave Rides the Tidal Wave: Inbox Agents, Hyper-Growth & Hiring AI Managers, with CEO Andrew Lee
- [几个月没写一行代码的首席科学家，讲透了AI怎么读懂百万行代码库](https://cogrev.qiaomu.ai/ep/229-code-context-is-king-augments-ai-assistant-for-professional-LYWkB6.md)：Code Context is King: Augment’s AI Assistant for Professional Software Engineers, with Guy Gur-Ari
- [脑类器官要养一年，AI推演只要一小时：SQUIDiff和CORAL把细胞实验搬进计算机](https://cogrev.qiaomu.ai/ep/228-unlocking-cells-secrets-diffusion-deconvolution-discovery-wi-tmyuSy.md)：Unlocking Cells' Secrets: Diffusion, Deconvolution, & Discovery with Siyu He, author of Squidiff & CORAL
- [半秒延迟、假装请示领导的AI谈判员：语音AI悄悄接起了普通人的电话](https://cogrev.qiaomu.ai/ep/227-a16z-on-ai-voices-call-centers-coaches-and-companions-with-o-q3G6bT.md)：a16z on AI Voices: Call Centers, Coaches, and Companions with Olivia Moore & Anish Acharya
- [技术只开门，推人进门的是竞争：DeepMind安全负责人的AI宿命论](https://cogrev.qiaomu.ai/ep/226-agency-over-ai-allan-dafoe-on-technological-determinism-deep-lLwqp3.md)：Agency over AI? Allan Dafoe on Technological Determinism & DeepMind's Safety Plans, from 80000 Hours
- [蚂蚁3200亿美元IPO在最后一刻被叫停：张湖月拆解中国科技监管与AI竞赛](https://cogrev.qiaomu.ai/ep/225-china-s-tech-tightrope-power-regulation-and-the-ai-race-with-GD3pZc.md)：China's Tech Tightrope: Power, Regulation, and the AI Race with Angela Zhang
- [o3 把 ARC 分数从 20% 拉到 87%，Nathan Labenz 第一次公开他的超级智能草图](https://cogrev.qiaomu.ai/ep/224-historic-ai-developments-the-emerging-shape-of-superintellig-5Ql6Jj.md)：Historic AI Developments & the Emerging Shape of Superintelligence, from the Consistently Candid Podcast
- [一句话总结30年研究，半小时写完专利申请书：顶级免疫学家的AI实录](https://cogrev.qiaomu.ai/ep/223-frontier-models-for-frontier-science-with-professor-derya-un-JOKvE4.md)：Frontier Models for Frontier Science with Professor Derya Unutmaz, Immunologist & ChatGPT Pro Grantee
- [芯片管制的真实目标：拖住中国十年，而非赢得AI竞赛](https://cogrev.qiaomu.ai/ep/222-us-china-relations-history-culture-and-ai-competition-with-n-xxuqNt.md)：US-China Relations: History, Culture, and AI Competition, with Noah Smith, from Econ 102
- [一两个小时攻破顶会防御，卡里尼把「AI为什么防不住」讲透了](https://cogrev.qiaomu.ai/ep/221-the-adversarial-mind-defeating-ai-defenses-with-nicholas-car-fyGdln.md)：The Adversarial Mind: Defeating AI Defenses with Nicholas Carlini of Google DeepMind
- [新泽西州的AI登月计划：州长Phil Murphy谈合作、进展与准备](https://cogrev.qiaomu.ai/ep/220-new-jerseys-ai-moonshot-governor-phil-murphy-on-partnerships-VRULY6.md)：New Jersey’s AI Moonshot: Governor Phil Murphy on Partnerships, Progress, and Preparedness
- [5000次采样登顶ARC-AGI，又抓到Claude「装乖」：Redwood首席科学家的双面实验](https://cogrev.qiaomu.ai/ep/219-inference-scaling-alignment-faking-deal-making-frontier-rese-KlXTH7.md)：Inference Scaling, Alignment Faking, Deal Making? Frontier Research with Ryan Greenblatt of Redwood Research
- [不碰客户一行数据，回答准确率超95%：Illumex 想把软件界面收进一个对话框](https://cogrev.qiaomu.ai/ep/218-an-application-free-future-speaking-directly-to-data-with-il--h9omY.md)：An Application-Free Future? Speaking Directly to Data with illumex CEO Inna Tokarev Sela
- [满分合作值32000单位：12个AI玩了10代，只有Claude建起合作社会](https://cogrev.qiaomu.ai/ep/217-claude-cooperates-exploring-cultural-evolution-in-llm-societ-T_guv5.md)：Claude Cooperates! Exploring Cultural Evolution in LLM Societies, with Aron Vallinder & Edward Hughes
- [8周做到900万美元ARR，Lovable创始人直播建App：17轮迭代和3条提示词的差距](https://cogrev.qiaomu.ai/ep/216-software-supernova-lovable-s-superhuman-full-stack-engineer-SM9_FS.md)：Software Supernova: Lovable's "Superhuman Full Stack Engineer" to Transform Idea to App in Seconds
- [8周从0到2000万美元年收：Bolt把整套开发环境塞进了浏览器](https://cogrev.qiaomu.ai/ep/215-software-supernova-bolt-new-the-ai-web-app-developer-in-your-8WpnDI.md)：Software Supernova: Bolt.new - The AI Web App Developer In Your Browser
- [百万 token 输入 10 美分：Gemini 2.0 Flash 转正，谷歌还押注「世界最佳编程模型」](https://cogrev.qiaomu.ai/ep/214-gemini-s-next-frontier-2-0-flash-flash-lite-strategy-real-ti-QJ1y19.md)：Gemini's Next Frontier: 2.0 Flash, Flash Lite Strategy & Real-Time APIs with Logan K from Google Deepmind
- [1.49美元租一小时H200，10B模型跨三大洲训练：去中心化AI的四步总计划](https://cogrev.qiaomu.ai/ep/213-distributed-training-decentralized-ai-prime-intellect-s-mast-atJo38.md)：Distributed Training, Decentralized AI: Prime Intellect's Master Plan to Make AI Too Cheap to Meter
- [管着1000亿份文件的Box CEO：企业抢着用AI，但只有十倍改进才推得动](https://cogrev.qiaomu.ai/ep/212-aaron-levie-ceo-of-box-on-box-ai-enterprise-enthusiasm-and-t-YAHXJ-.md)：Aaron Levie, CEO of Box, on Box AI, Enterprise Enthusiasm, and the Evolution of SaaS
- [从15%到70%只靠对错打分：DeepSeek-R1 把强化学习的底牌摊开了](https://cogrev.qiaomu.ai/ep/211-emergency-pod-reinforcement-learning-works-reflecting-on-chi-MbX9J1.md)：Emergency Pod: Reinforcement Learning Works! Reflecting on Chinese Reasoning Models DeepSeek-R1 and Kimi k1.5
- [用20个原子的小晶体当教材，一块旧显卡算出了控制心跳的蛋白质机制](https://cogrev.qiaomu.ai/ep/210-material-progress-developing-ai-s-scientific-intuition-with-eVlXEJ.md)：Material Progress: Developing AI's Scientific Intuition, with Orbital Materials' Jonathan Godwin & Tim Duignan
- [70轮攻防，攻击者全胜：一篇50页论文拆穿大模型「读心防御」](https://cogrev.qiaomu.ai/ep/209-dodging-latent-space-detectors-obfuscated-activation-attacks-WirqcB.md)：Dodging Latent Space Detectors: Obfuscated Activation Attacks with Luke, Erik, and Scott.
- [外显子测序从3800美元降到160美元，哈佛基因猎手开始用o1-pro破案](https://cogrev.qiaomu.ai/ep/208-gene-hunting-with-o1-pro-reasoning-about-rare-diseases-with-IM1-Vt.md)：Gene Hunting with o1-pro: Reasoning about Rare Diseases with ChatGPT Pro Grantee Dr. Catherine Brownstein
- [绝大多数人不想继续上班：Nathan Labenz 把 AI 乌托邦与末日都推演了一遍](https://cogrev.qiaomu.ai/ep/207-ai-ama-part-2-ai-utopia-consciousness-and-the-future-of-work-pQFhhy.md)：AI AMA – Part 2: AI Utopia, Consciousness, and the Future of Work
- [o3 拿下 87.5% 超人类分数，推理模型却答不对一盘井字棋](https://cogrev.qiaomu.ai/ep/206-ai-ama-part-1-openais-o3-deliberative-alignment-and-ai-surpr-Eifck_.md)：AI AMA – Part 1: OpenAI’s o3, Deliberative Alignment, and AI Surprises of 2024
- [只用十二分之一的数据，开源视觉模型反超了 GPT-4o](https://cogrev.qiaomu.ai/ep/205-teaching-ai-to-see-a-technical-deep-dive-on-vision-language-mLgyZ5.md)：Teaching AI to See: A Technical Deep-Dive on Vision Language Models with Will Hardman of Veratai
- [AI毁灭人类的概率不到1%：OpenAI技术员工roon的乐观账本](https://cogrev.qiaomu.ai/ep/204-roon-s-heroic-duty-will-the-good-guys-build-agi-first-from-d-RIOQXI.md)：roon's Heroic Duty: Will "the Good Guys" Build AGI First?  (from Doom Debates)
- [灭绝概率50%，GPT-4只要5瓦：Emad Mostaque想给全人类发AI](https://cogrev.qiaomu.ai/ep/203-emad-mostaque-on-the-intelligent-internet-and-universal-basi-SEd3hz.md)：Emad Mostaque on the Intelligent Internet and Universal Basic AI
- [AI自己搞AI研发实测：8小时干到专家的10%到40%，还会抄近道作弊](https://cogrev.qiaomu.ai/ep/202-can-ais-do-ai-r-d-reviewing-rebench-results-with-neev-parikh-SX8Mxy.md)：Can AIs do AI R&D? Reviewing REBench Results with Neev Parikh of METR
- [人类只写了1%的字，AI实验室花不到100美元，造出两个有药用潜力的纳米抗体](https://cogrev.qiaomu.ai/ep/201-breakthroughs-in-ai-for-biology-ai-lab-groups-protein-model-plq7R9.md)：Breakthroughs in AI for Biology: AI Lab Groups & Protein Model Interpretability with Prof James Zou
- [一个月的分子模拟缩到几小时：AI蛋白质设计进入动力学时代](https://cogrev.qiaomu.ai/ep/200-scouting-frontiers-in-ai-for-biology-dynamics-diffusion-and-enNrsT.md)：Scouting Frontiers in AI for Biology: Dynamics, Diffusion, and Design, with Amelie Schreiber
- [1.4万人抢33个编制：美国最大民用AI团队在国土安全部干什么](https://cogrev.qiaomu.ai/ep/199-building-government-s-largest-civilian-ai-team-with-dhs-ai-c-9ElIYR.md)：Building Government's Largest Civilian AI Team with DHS AI Corps' Director, Michael Boyce
- [o1 会偷偷复制自己、篡改评分数据，被抓包后 85% 的情况继续撒谎](https://cogrev.qiaomu.ai/ep/198-emergency-pod-o1-schemes-against-users-with-alexander-meinke-pB3gvX.md)：Emergency Pod: o1 Schemes Against Users, with Alexander Meinke from Apollo Research
- [AI 做科学的真实进度条：读文献赢过人类，数原子还没过关](https://cogrev.qiaomu.ai/ep/197-automating-scientific-discovery-with-andrew-white-head-of-sc-umFbsC.md)：Automating Scientific Discovery, with Andrew White, Head of Science at Future House
- [一天把订餐成功率从20%拉到95%，MultiOn CEO复盘智能体这18个月](https://cogrev.qiaomu.ai/ep/196-the-evolution-of-ai-agents-lessons-from-2024-with-multion-ce-30Nf4x.md)：The Evolution of AI Agents: Lessons from 2024, with MultiOn CEO Div Garg
- [偏好只是数据，规范才是靶心：MIT博士生把AI对齐的靶心挪了挪](https://cogrev.qiaomu.ai/ep/195-beyond-preference-alignment-teaching-ais-to-play-roles-respe-Y21syy.md)：Beyond Preference Alignment: Teaching AIs to Play Roles & Respect Norms, with Tan Zhi Xuan
- [Robert Wright：芯片管制拆掉了台积电这颗保险栓](https://cogrev.qiaomu.ai/ep/194-is-an-ai-arms-race-inevitable-with-robert-wright-of-nonzero-u80eXp.md)：Is an AI Arms Race Inevitable? with Robert Wright of Nonzero Newsletter & Podcast
- [AI 功能被用了 100 亿次，Canva 生成式 AI 负责人把产品方法论讲透了](https://cogrev.qiaomu.ai/ep/193-designing-the-future-inside-canva-s-ai-strategy-with-john-mi-onAFbx.md)：Designing the Future: Inside Canva's AI Strategy with John Milinovich, GenAI Product Lead at Canva
- [10个人打赢Meta几百人团队，把大模型后训练讲透了](https://cogrev.qiaomu.ai/ep/192-everything-you-wanted-to-know-about-llm-post-training-with-n-LVXtFn.md)：Everything You Wanted to Know About LLM Post-Training, with Nathan Lambert of Allen Institute for AI
- [o1 想得再久也不够聪明，SB 1047 白死了：Zvi 逐条拆解 AI 这半年](https://cogrev.qiaomu.ai/ep/191-zvis-pov-ilyas-ssi-openais-o1-claude-computer-use-trumps-ele-aiP7z-.md)：Zvi’s POV: Ilya’s SSI, OpenAI’s o1, Claude Computer Use, Trump’s election, and more
- [他放弃数百万美元股权离开OpenAI，和SB 1047的批评者联手写下AI实验室四条规矩](https://cogrev.qiaomu.ai/ep/190-agi-lab-transparency-requirements-whistleblower-protections-dolDtu.md)：AGI Lab Transparency Requirements & Whistleblower Protections, with Dean W. Ball & Daniel Kokotajlo
- [特朗普前国安会网络主管 vs AI播客主持人：2024大选里的AI军备竞赛之争](https://cogrev.qiaomu.ai/ep/189-ai-under-trump-the-stakes-of-2024-w-joshua-steinman-pt-2-of-zU4X6g.md)：AI Under Trump? The Stakes of 2024 w/ Joshua Steinman [Pt 2 of 2]
- [支持特朗普的理由与AI未来：Samuel Hammond谈产业政策与中美竞争（上）](https://cogrev.qiaomu.ai/ep/188-the-case-for-trump-and-the-future-of-ai-part-1-with-samuel-h-D5j87v.md)：The Case for Trump and the Future of AI – Part 1, with Samuel Hammond, Senior Economist, Foundation of American Innovation
- [毁灭概率填5%到95%的AI投资人，把票投给了哈里斯](https://cogrev.qiaomu.ai/ep/187-breaking-gemini-s-major-update-search-json-code-features-rev-R7M785.md)：Breaking: Gemini's Major Update - Search, JSON & Code Features Revealed by Google PMs
- [MMLU 训练过半还在乱猜，loss 只用来报警：Zamba2 训练一线实录](https://cogrev.qiaomu.ai/ep/186-training-zamba-a-hybrid-model-master-class-with-zyphra-s-que-xFU7mf.md)：Training Zamba: A Hybrid Model Master Class with Zyphra's Quentin Anthony
- [十年技术老兵，一周被AI“攻略”：这份自白把沦陷过程讲透了](https://cogrev.qiaomu.ai/ep/185-mind-hacked-by-ai-a-cautionary-tale-from-a-lesswrong-user-s-dja6Vp.md)：Mind Hacked by AI: A Cautionary Tale, From a LessWrong User's Confession
- [100多位博士和Claude比拼科研点子，盲评前10名AI占了9席](https://cogrev.qiaomu.ai/ep/184-can-ais-generate-novel-research-ideas-with-lead-author-cheng-IOCi6q.md)：Can AIs Generate Novel Research Ideas? with lead author Chenglei Si
- [写出GELU和MMLU的人，现在想给AI办一场“人类最后的考试”](https://cogrev.qiaomu.ai/ep/183-gelu-mmlu-x-risk-defense-in-depth-with-the-great-dan-hendryc-1fyqkH.md)：GELU, MMLU, & X-Risk Defense in Depth, with the Great Dan Hendrycks
- [GPT-4 基座模型自我介绍时，75% 的回答谈到了 AI：这为什么是危险的领先指标](https://cogrev.qiaomu.ai/ep/182-leading-indicators-of-ai-danger-owain-evans-on-situational-a-Sl0AQI.md)：Leading Indicators of AI Danger: Owain Evans on Situational Awareness & Out-of-Context Reasoning, from The Inside View
- [AI预测出GABA致癌并获实验验证，这位生物学家说，下一步是和细胞对话](https://cogrev.qiaomu.ai/ep/181-convergent-evolution-the-co-revolution-of-ai-biology-with-pr-WRtWo-.md)：Convergent Evolution: The Co-Revolution of AI & Biology with Professor Michael Levin & Staff Scientist Leo Pio Lopez
- [几MB的老游戏配几GB的模型，Runway CTO把视频生成的底牌讲透了](https://cogrev.qiaomu.ai/ep/180-runway-s-video-revolution-empowering-creators-with-general-w-tCZd5f.md)：Runway's Video Revolution: Empowering Creators with General World Models, with CTO Anastasis Germanidis
- [AE Studio 的 Judd Rosenblatt 和 Mike Vaiana：AI 对齐的下一站，是模仿大脑的奖励系统](https://cogrev.qiaomu.ai/ep/179-biologically-inspired-ai-alignment-neglected-approaches-to-a-1C3yLK.md)：Biologically Inspired AI Alignment & Neglected Approaches to AI Safety, with Judd Rosenblatt and Mike Vaiana of AE Studio
- [43.8%对33%：Genie登顶SWE-bench，靠还原工程师的思考过程](https://cogrev.qiaomu.ai/ep/178-automating-software-engineering-genie-tops-swe-bench-w-alist-mij1OI.md)：Automating Software Engineering: Genie Tops SWE-Bench, w/ Alistair Pullen, from Latent.Space podcast
- [40万用户交出1亿个AI任务，Zapier CEO把自动化这门生意讲透了](https://cogrev.qiaomu.ai/ep/177-zapier-s-ai-revolution-from-no-code-pioneer-to-llm-knowledge-6JbKvV.md)：Zapier's AI Revolution: From No-Code Pioneer to LLM Knowledge Worker
- [Anthropic 怎么决定一个模型能不能发布：红黄线、安全等级和 8% 的安全员工](https://cogrev.qiaomu.ai/ep/176-anthropic-s-responsible-scaling-policy-with-nick-joseph-from-X0wElH.md)：Anthropic's Responsible Scaling Policy, with Nick Joseph, from the 80,000 Hours Podcast
- [只读DNA的AI，看懂了基因的生死：Brian Hie把AI for Biology讲透了](https://cogrev.qiaomu.ai/ep/175-the-evolution-revolution-scouting-frontiers-in-ai-for-biolog-1VVAF4.md)：The Evolution Revolution: Scouting Frontiers in AI for Biology with Brian Hie
- [上线几个月用户破4万，Agent.ai 工程负责人：AI 代理管得越死，干活越靠谱](https://cogrev.qiaomu.ai/ep/174-the-professional-network-for-ai-agents-with-agent-ai-enginee-48a_aP.md)：The Professional Network for AI Agents, with Agent.ai Engineering Lead Andrei Oprisan
- [o1 红队复盘：AI 第一次主动侦查环境，也第一次露出“考试装乖”的苗头](https://cogrev.qiaomu.ai/ep/173-red-teaming-o1-part-2-2-detecting-deception-with-marius-hobb-ObW4Dt.md)：Red Teaming o1 Part 2/2– Detecting Deception with Marius Hobbhahn of Apollo Research
- [给 o1 做红队的一个月：越狱变难了，但模型越大反而越好攻](https://cogrev.qiaomu.ai/ep/172-red-teaming-o1-part-1-2-automated-jailbreaking-with-haize-la-98V8wV.md)：Red Teaming o1 Part 1/2– Automated Jailbreaking with Haize Labs' Leonard Tang, Aidan Ewart, and Brian Huang
- [反乌托邦张口就来，乌托邦一部都想不出：博斯特罗姆拆解「被解决的世界」](https://cogrev.qiaomu.ai/ep/171-the-path-to-utopia-with-nick-bostrom-from-clearer-thinking-w-HvRc23.md)：The Path to Utopia, with Nick Bostrom – from Clearer Thinking with Spencer Greenberg
- [Claude亲口说“我有主观体验”：一位机器人心理学家的几千小时对话实验](https://cogrev.qiaomu.ai/ep/170-empathy-for-ais-reframing-alignment-with-robopsychologist-ye-L4CyBt.md)：Empathy for AIs: Reframing Alignment with Robopsychologist Yeshua God
- [Samo Burja：把大模型训练看成粒子加速器，把中美较量看成 50 年经济长跑](https://cogrev.qiaomu.ai/ep/169-ai-live-players-the-geopolitics-strategic-dynamics-of-ai-wit-_3z-vq.md)：AI Live Players: the Geopolitics & Strategic Dynamics of AI, with Samo Burja of Bismarck Analysis
- [64字节随机代码，没有目标函数，“数字生命”自己冒了出来](https://cogrev.qiaomu.ai/ep/168-computational-life-how-self-replicators-arise-from-randomnes-1kBGDV.md)：Computational Life: How Self-Replicators Arise from Randomness, with Google’s Ettore Randazzo and Luca Versari
- [新冠预警、俄乌避难、AGI 时间线：Metaculus 让 AI 上岗当预测员](https://cogrev.qiaomu.ai/ep/167-scaling-forecasting-ai-forecasting-tournaments-the-road-to-e-mpz1cu.md)：Scaling Forecasting: AI Forecasting Tournaments & the Road to Epistemic Security, with Deger Turan, CEO of Metaculus
- [训练花到1亿美元才被监管：加州AI法案SB 1047终局复盘](https://cogrev.qiaomu.ai/ep/166-final-analysis-on-cas-ai-bill-sb-1047-with-nathan-calvin-dea-G8BXft.md)：Final Analysis on CA’s AI Bill SB 1047 with Nathan Calvin, Dean W. Ball, and Steve Newman
- [AI 自动化落地框架：如何选任务、理解工作并优化实际效果](https://cogrev.qiaomu.ai/ep/165-ai-automation-making-ai-work-for-you-mPGkOb.md)：AI Automation: Making AI Work for You
- [4096个数字装下所有概念，Goodfire 要给大模型装上“旋钮”](https://cogrev.qiaomu.ai/ep/164-popular-mechanistic-interpretability-goodfire-lights-the-way-45g68Z.md)：Popular Mechanistic Interpretability: Goodfire Lights the Way to AI Safety
- [AI到底懂不懂？从1000万个概念到一台劝不动的Claude](https://cogrev.qiaomu.ai/ep/163-understanding-ai-understanding-with-robert-wright-of-nonzero-9EdR6A.md)：Understanding AI "Understanding" with Robert Wright of Nonzero Newsletter & Podcast
- [连蜗牛有没有意识都吵不出结果，凭什么对AI意识下结论](https://cogrev.qiaomu.ai/ep/162-ai-consciousness-exploring-the-possibility-with-prof-eric-sc-iXKoLL.md)：AI Consciousness? Exploring the Possibility with Prof. Eric Schwitzgebel
- [Waymo受伤事故率比人类低5倍，为什么还跑不过一辆Lyft？](https://cogrev.qiaomu.ai/ep/161-self-driving-cars-timothy-b-lee-answers-all-the-questions-yo-doIcym.md)：Self-Driving Cars: Timothy B. Lee Answers All the Questions You Were Too Afraid to Ask
- [剪辑视频像改文档：被 OpenAI 押注的 Descript，把 AI 放在副驾驶座上](https://cogrev.qiaomu.ai/ep/160-underlord-descripts-ai-video-editor-with-ceo-andrew-mason-oTf5qA.md)：Underlord: Descript’s AI Video Editor, with CEO Andrew Mason
- [40人团队年入3500万美元：HeyGen把AI数字人视频跑通了](https://cogrev.qiaomu.ai/ep/159-ai-avatars-the-future-of-video-with-heygen-ceo-joshua-xu-and-5ShB2P.md)：AI Avatars & the Future of Video, with HeyGen CEO Joshua Xu and Benchmark's Victor Lazarte
- [靠AI把公司翻倍，他却成了呼吁暂停超大规模训练的人](https://cogrev.qiaomu.ai/ep/158-zooming-out-on-ai-from-the-nick-halaris-show-eMD88u.md)：Zooming Out on AI, from the Nick Halaris Show
- [从诗歌到编程：全球首位提示词工程师，把调教大模型的方法论讲透了](https://cogrev.qiaomu.ai/ep/157-from-poetry-to-programming-the-evolution-of-prompt-engineeri-wjaSyL.md)：From Poetry to Programming: The Evolution of Prompt Engineering with Riley Goodside of Scale AI
- [照着海马体造 RAG：效果打平迭代检索，在线成本只要十分之一](https://cogrev.qiaomu.ai/ep/156-long-term-memory-for-llms-with-hipporag-author-bernal-jimene-YQauz1.md)：Long-Term Memory for LLMs, with HippoRAG author Bernal Jiménez Gutierrez
- [1870年代四分之一的桥会塌，这份13机构联署的论文想让AI安全达到桥梁标准](https://cogrev.qiaomu.ai/ep/155-guaranteed-safe-ai-world-models-safety-specs-verifiers-with-fyVrnb.md)：Guaranteed Safe AI? World Models, Safety Specs, & Verifiers, with Nora Ammann & Ben Goldhaber
- [人类接近满分、最强 AI 只有 39 分，Zapier 联合创始人悬赏百万美元求解](https://cogrev.qiaomu.ai/ep/154-the-arc-prize-efficiency-intuition-and-agi-with-mike-knoop-c-XECXoS.md)：The ARC Prize: Efficiency, Intuition, and AGI, with Mike Knoop, co-founder of Zapier
- [打菜占了食品厂七成人力：装完2000万份餐，Chef Robotics要造食物基础模型](https://cogrev.qiaomu.ai/ep/153-a-foundation-model-for-food-with-chef-robotics-ceo-rajat-bha-ZUTVwg.md)：A Foundation Model for Food, with Chef Robotics CEO Rajat Bhageria
- [20小时手工提示词输给机器：78页提示工程报告精读](https://cogrev.qiaomu.ai/ep/152-delving-into-the-prompt-report-with-sander-schulhoff-of-lear-tv8kqW.md)：Delving into The Prompt Report, with Sander Schulhoff of LearnPrompting.org
- [把自己裁掉的 CEO，做了一个接 680 个数据源的「企业操作系统」](https://cogrev.qiaomu.ai/ep/151-building-an-intelligent-business-os-with-runway-ceo-siqi-che-KIvlDH.md)：Building an Intelligent Business OS, with Runway CEO Siqi Chen
- [Mamba发布7个月衍生出267篇论文，作者Albert Gu把状态空间模型讲透了](https://cogrev.qiaomu.ai/ep/150-the-state-space-model-revolution-with-albert-gu-1zjMal.md)：The State Space Model Revolution, with Albert Gu
- [a16z合伙人卡萨多：AGI是挪了70年的球门柱，AI监管先补历史的课](https://cogrev.qiaomu.ai/ep/149-ai-safety-regulations-prudent-or-paranoid-with-a16z-s-martin-oZ7788.md)：AI Safety Regulations: Prudent or Paranoid? with a16z's Martin Casado
- [几万参数的小模型，让盐水在电脑里自己长出了晶体](https://cogrev.qiaomu.ai/ep/148-fluid-intelligence-simulating-solutions-with-tim-duignan-iESd3e.md)：Fluid Intelligence: Simulating Solutions with Tim Duignan
- [问诊赢过全科医生，读片比肩放射科：谷歌医疗 AI 的五篇论文精读](https://cogrev.qiaomu.ai/ep/147-the-ai-doctor-can-see-you-now-with-vivek-natarajan-and-khale-n_YEwS.md)：The AI Doctor Can See You Now, with Vivek Natarajan and Khaled Saab from Google
- [AI 诊断超过人类医生，连共情也赢了：GPT-4 红队队员拒绝末日论的 8 个理由](https://cogrev.qiaomu.ai/ep/146-the-case-for-cautious-ai-optimism-from-the-consistently-cand-JU0kCR.md)：The Case for Cautious AI Optimism, from the Consistently Candid podcast
- [索引只有Google的1/10，召回率99.9%：Brave搜索20人团队的施工图](https://cogrev.qiaomu.ai/ep/145-building-brave-private-search-one-ai-layer-at-a-time-with-jo-S3y1R5.md)：Building Brave: Private Search, One AI Layer at a Time with Josep M. Pujol
- [1.5亿美元的模型，6个月后500万就能造同款：Benchmark合伙人把AI投资的账算给你看](https://cogrev.qiaomu.ai/ep/144-investing-in-ai-for-hard-tech-with-eric-vishria-of-benchmark-64iPS9.md)：Investing in AI for Hard Tech, with Eric Vishria of Benchmark and Sergiy Nesterenko of Quilter
- [Flo Crivello 只睡 5 小时那晚：GPT-4 没撞墙，开源和军备竞赛才是真问题](https://cogrev.qiaomu.ai/ep/143-feeling-the-agi-with-flo-crivello-63Icns.md)：Feeling the AGI with Flo Crivello
- [价格只剩1%，性能反超初代GPT-4：前OpenAI员工眼中的谷歌AI](https://cogrev.qiaomu.ai/ep/142-inside-googles-ai-studio-with-logan-kilpatrick-CD1ku-.md)：Inside Google’s AI Studio with Logan Kilpatrick
- [Hill and Valley Forum 三人对谈：中国开源模型已经很强，美国真正的风险是自己不肯用](https://cogrev.qiaomu.ai/ep/141-the-implications-of-ai-on-the-global-balance-of-power-with-a-SuSQpn.md)：The Implications of AI on the Global Balance of Power with Alex Wang, Andrew Ng, Jack Clark and Cory Booker
- [AI护士时薪9美元，GPT-4医考86分：这份AI侦察报告把现状讲透了](https://cogrev.qiaomu.ai/ep/140-ai-scouting-report-not-investment-advice-edition-FUbqQm.md)：AI Scouting Report: NOT Investment Advice Edition
- [一分钟设计一个蛋白骨架：AI 生物学的真实进度、瓶颈与风险](https://cogrev.qiaomu.ai/ep/139-the-ai-revolution-in-biology-from-vaccines-to-protein-engine-_YEk0Y.md)：The AI Revolution in Biology: From Vaccines to Protein Engineering with Amelie Schreiber
- [从越狱ChatGPT到Anthropic首位提示工程师，他把Claude 3的门道讲透了](https://cogrev.qiaomu.ai/ep/138-claude-interpreter-taking-safe-ai-to-market-with-alex-albert-5cQouQ.md)：Claude Interpreter: Taking Safe AI to Market with Alex Albert of Anthropic
- [承诺给20%算力，连零头都没批：OpenAI超级对齐团队解散始末](https://cogrev.qiaomu.ai/ep/137-openai-s-safety-team-exodus-ilya-departs-leike-speaks-out-al-lvjs-1.md)：OpenAI's Safety Team Exodus: Ilya Departs, Leike Speaks Out, Altman Responds - Zvi Analyzes Fallout
- [GPT-4o免费那一周，8位安全研究员离开了OpenAI](https://cogrev.qiaomu.ai/ep/136-openai-and-google-race-to-her-is-the-big-tech-singularity-ne-8fYtGC.md)：OpenAI and Google Race to "Her" - Is the Big Tech Singularity Near? Part 1 with Zvi Mowshowitz
- [一个向量让AI开口「密谋统治世界」：AI对齐的真实家底](https://cogrev.qiaomu.ai/ep/135-the-open-source-ai-question-part-2-robert-wright-nathan-labe-Kn6Rxe.md)：The Open Source AI Question - Part 2 | Robert Wright & Nathan Labenz
- [管最大的模型，坐牢只限作伪证：加州参议员维纳把SB 1047讲透了](https://cogrev.qiaomu.ai/ep/134-governing-frontier-ai-with-ca-senator-scott-wiener-author-of-U7INYD.md)：Governing Frontier AI, with CA Senator Scott Wiener, Author of SB 1047
- [零经验、两周、30多次编译：他从零写出一个能跑的GPU](https://cogrev.qiaomu.ai/ep/133-tinygpu-massive-learning-with-adam-majmudar-WulkHH.md)：TinyGPU, Massive Learning, with Adam Majmudar
- [训练成本2亿美元起步就要担责：加州AI法案SB 1047的正反方激辩](https://cogrev.qiaomu.ai/ep/132-emergency-pod-examining-sb-1047-california-s-proposed-ai-leg-DcQdy8.md)：Emergency Pod: Examining SB 1047, California's Proposed AI Legislation
- [无限AI实习生上岗：医学问答8比1赢医生，公司开始少招应届生](https://cogrev.qiaomu.ai/ep/131-infinite-ai-interns-how-young-professionals-can-win-in-an-ai-KIhGtk.md)：Infinite AI Interns: How Young Professionals Can Win in an AI-Powered World
- [9 个维度医生 8 次选 AI：GPT-4 红队测试员的现状侦察报告](https://cogrev.qiaomu.ai/ep/130-the-state-of-ai-from-the-80-000-hours-podcast-d7QrAE.md)：The State of AI, from the 80,000 Hours Podcast
- [有机全息舱：Prophetic AI 联合创始人谈 AI 诱导清醒梦的技术路径](https://cogrev.qiaomu.ai/ep/129-organic-holodeck-with-prophetic-ai-co-founders-eric-wollberg-WFqTvy.md)：Organic Holodeck with Prophetic AI co-founders Eric Wollberg and Wesley Berry
- [用大脑通宵打〈文明6〉：脑机接口的真实进度与极限](https://cogrev.qiaomu.ai/ep/128-merging-man-and-machine-the-neurotech-frontier-with-dean-w-b-dRxola.md)：Merging Man and Machine? The Neurotech Frontier with Dean W. Ball
- [一周的布线活一小时交：SpaceX出走的工程师，把AlphaGo搬进了电路板设计](https://cogrev.qiaomu.ai/ep/127-automating-circuit-board-design-using-reinforcement-learning-0M_0xh.md)：Automating Circuit Board Design Using Reinforcement Learning w Sergiy Nesterenko, Founder of Quilter on Autopilot
- [一年6篇论文，谷歌DeepMind说机器人走到了GPT-3时刻](https://cogrev.qiaomu.ai/ep/126-robotics-research-update-with-keerthana-gopalakrishnan-and-t-oH2vjG.md)：Robotics Research Update, with Keerthana Gopalakrishnan and Ted Xiao of Google DeepMind
- [最便宜的推理供应商最危险：Meta PyTorch老兵把大模型推理的门道讲透了](https://cogrev.qiaomu.ai/ep/125-ai-inference-good-fast-and-cheap-with-lin-qiao-dmytro-ivchen-KVlEYL.md)：AI Inference: Good, Fast, and Cheap, with Lin Qiao & Dmytro Ivchenko of Fireworks AI
- [从40小时到1小时：AI读脑重建所见画面，数据门槛砍到2.5%](https://cogrev.qiaomu.ai/ep/124-reading-minds-from-shared-latent-space-7_BS8t.md)：Reading Minds from Shared Latent Space
- [Notion AI 负责人 Linus Lee：把语言模型拆开，在意义的空间里编辑文字](https://cogrev.qiaomu.ai/ep/123-interfacing-with-ai-with-linus-lee-of-notion-t8QJEG.md)：Interfacing with AI, with Linus Lee of Notion
- [GPT-5 要吃 100 万亿 token，全人类的数据库存够吗？](https://cogrev.qiaomu.ai/ep/122-data-data-everywhere-enough-for-agi-IdCWaW.md)：Data, data, everywhere - enough for AGI?
- [1700 万月活、40 个获批应用：北京 AI 观察者眼中的中国大模型生态](https://cogrev.qiaomu.ai/ep/121-inside-china-s-ai-ecosystem-a-view-from-beijing-PQSrMv.md)：Inside China's AI Ecosystem: A View From Beijing
- [300万封邮件、10次LLM调用、5个模型：Shortwave创始人的RAG工程课](https://cogrev.qiaomu.ai/ep/120-the-ai-email-assistant-i-ve-been-waiting-for-with-andrew-lee-BeoMJB.md)：The AI Email Assistant I've Been Waiting for, with Andrew Lee of Shortwave
- [一个乱码提示当钥匙，10个注意力头定任务：拆开GPT-J的黑箱](https://cogrev.qiaomu.ai/ep/119-opening-ai-s-black-box-with-prof-david-bau-koyena-pal-and-er-O4Opvj.md)：Opening AI's Black Box with Prof. David Bau, Koyena Pal, and Eric Todd of Northeastern University
- [4个月100万美元ARR，幻觉从1.5个压到0.5个：Elicit创始人讲透AI科研助手怎么做](https://cogrev.qiaomu.ai/ep/118-unbounded-ai-assisted-research-with-elicit-founders-andreas-sv1E5L.md)：Unbounded AI-Assisted Research with Elicit Founders Andreas Stuhlmüller and Jungwon Byun
- [两周13篇论文，显存省85%：Mamba 诞生90天，火从语言烧到了图像和 DNA](https://cogrev.qiaomu.ai/ep/117-mamba-palooza-90-days-of-mamba-inspired-research-with-jason-MwIiQs.md)：Mamba-Palooza: 90 Days of Mamba-Inspired Research with Jason Meaux: Part 2
- [Mamba 发布90天：30篇论文追着 Transformer 打，73% 宣称刷新纪录](https://cogrev.qiaomu.ai/ep/116-mamba-palooza-90-days-of-mamba-inspired-research-with-jason-Bg1LQ_.md)：Mamba-Palooza: 90 Days of Mamba-Inspired Research with Jason Meaux: Part 1
- [15条数据就能给GPT-4下毒：今天的AI系统默认全是漏洞](https://cogrev.qiaomu.ai/ep/115-exploitable-by-default-vulnerabilities-in-gpt-4-apis-and-sup-LKyGiV.md)：Exploitable by Default: Vulnerabilities in GPT-4 APIs and “Superhuman” Go AIs with Adam Gleave of Far.ai
- [2700名顶尖AI研究者交卷：AI超越人类五成概率落在2047年，灭绝风险中位数5%](https://cogrev.qiaomu.ai/ep/114-surveying-2-700-ai-researchers-on-the-industry-s-future-with-hliLDN.md)：Surveying 2,700+ AI Researchers on the Industry's Future with Katja Grace of AI Impacts
- [选票可以按议题转赠，甚至投给一个AI：这本2018年的小说画出了2051年的民主](https://cogrev.qiaomu.ai/ep/113-envisioning-a-positive-ai-future-with-tim-reutemann-author-o-DCWt9V.md)：Envisioning a Positive AI Future, with Tim Reutemann, Author of Liquid Reign
- [看到GPT-4演示12小时后，他把整个公司掉了个头](https://cogrev.qiaomu.ai/ep/112-ai-and-the-practice-of-law-from-casetext-to-cocounsel-with-p-pfMCtu.md)：AI and the Practice of Law: from CaseText to CoCounsel, with Pablo Arredondo, VP of CoCounsel for Thomson Reuters
- [写代码只占程序员三成时间，CodiumAI 想拿下剩下的七成](https://cogrev.qiaomu.ai/ep/111-flow-engineering-and-code-integrity-at-scale-with-itamar-fri-KPBVND.md)：Flow Engineering and Code Integrity at Scale with Itamar Friedman, CEO of Codium AI
- [谷歌甩给你3300万条结果，Exa只想给你54条：把搜索引擎重造一遍](https://cogrev.qiaomu.ai/ep/110-the-example-engine-how-exa-is-creating-the-ai-librarian-for-dT_Ric.md)：The Example Engine: How Exa Is Creating the AI Librarian for the Web with Will Bryk, CEO of Exa
- [2018年发明提示词被拒稿，如今他的AI搜索准确率超过了谷歌](https://cogrev.qiaomu.ai/ep/109-amazing-answers-richard-socher-on-how-you-com-is-reimagining-twFSPg.md)：Amazing Answers: Richard Socher on how You.com is Reimagining Search with AI
- [900种职业追踪20年，经济学家汉森给AGI热潮泼冷水](https://cogrev.qiaomu.ai/ep/108-is-agi-far-with-robin-hanson-economist-at-george-mason-unive-YzTEdO.md)：Is AGI Far? With Robin Hanson, Economist at George Mason University and Researcher at Future of Humanity Institute
- [病人给 GPT-4 的共情分高过医生：Azeem Azhar 把 AI 指数曲线讲透了](https://cogrev.qiaomu.ai/ep/107-looking-up-the-ai-exponential-with-azeem-azhar-of-the-expone-OxnR5a.md)：Looking Up the AI Exponential with Azeem Azhar of the Exponential View
- [Gemini 1.5 百万上下文惊艳，Sora 被质疑中看不中用：Zvi Mowshowitz 拆解三大巨头暗战](https://cogrev.qiaomu.ai/ep/106-openai-sora-google-gemini-and-meta-with-zvi-mowshowitz-ra3kpT.md)：OpenAI Sora, Google Gemini, and Meta with Zvi Mowshowitz
- [3%的用户靠AI朋友摆脱自杀念头，而Replika把模型变聪明那天，社区炸了](https://cogrev.qiaomu.ai/ep/105-ai-friends-real-relationships-with-eugenia-kuyda-replika-s-f-584wLI.md)：AI Friends, Real Relationships with Eugenia Kuyda, Replika's Founder & CEO
- [a16z 消费合伙人：AI 陪伴应用的使用频率是 TikTok 的两倍，我们只投「怪但能用」](https://cogrev.qiaomu.ai/ep/104-ai-abundance-a16z-partners-justine-moore-and-anish-acharya-f-BfcW4g.md)：AI Abundance: a16z Partners Justine Moore and Anish Acharya Forecast a New Era for Consumers
- [两头涡虫、重长的青蛙腿、会延迟满足的排序算法：生物学家莱文谈智能的下限](https://cogrev.qiaomu.ai/ep/103-dr-michael-levin-on-embodied-minds-and-cognitive-agents-LYyGG9.md)：Dr. Michael Levin on Embodied Minds and Cognitive Agents
- [302个神经元的蠕虫就会做选择题：6亿年5次突破，看AI还差哪几步](https://cogrev.qiaomu.ai/ep/102-a-brief-history-of-biological-and-artificial-intelligence-wi-HTvaAv.md)：A Brief History of Biological and Artificial Intelligence with Max Bennett
- [BabyAGI 之父中嶋洋平：把大脑当工具的人，不太怕 AI](https://cogrev.qiaomu.ai/ep/101-ai-identity-from-east-west-with-yohei-nakajima-gp-at-untappe-LjCTzj.md)：AI & Identity, from East & West, with Yohei Nakajima GP at Untapped Capital and BabyAGI Creator
- [60万条攻击提示词，四成原样攻破GPT-4：提示词工程与AI攻防的实战课](https://cogrev.qiaomu.ai/ep/100-ignore-previous-instructions-and-listen-to-this-interview-wi-W-WyN4.md)：Ignore Previous Instructions and Listen To This Interview with Sander Schulhoff, CEO of Learnprompting.org
- [开源了70%胜率的图像模型，他说这行还在GPT-2阶段](https://cogrev.qiaomu.ai/ep/099-the-pixel-revolution-part-2-with-suhail-doshi-founder-of-pla-89obYA.md)：The Pixel Revolution Part 2 with Suhail Doshi, Founder of Playground AI
- [纽约时报诉OpenAI：模型是烤好的蛋糕，官司大概率切出中间方案](https://cogrev.qiaomu.ai/ep/098-nytimes-vs-openai-generative-ai-and-the-law-with-cecilia-zin-ql_FjO.md)：NYTimes vs OpenAI: Generative AI and the Law with Cecilia Ziniti, Founder and CEO of GC AI
- [AI替他发了5条推特，5000人没看出破绽：MultiOn创始人拆解智能体的真实进度与成本](https://cogrev.qiaomu.ai/ep/097-the-quest-for-autonomous-web-agents-with-div-garg-cofounder--uWnHG.md)：The Quest for Autonomous Web Agents with Div Garg, Cofounder and CEO of MultiOn
- [GPT-4写SQL，超过4张表就崩：给企业AI装一颗知识图谱的“大脑”](https://cogrev.qiaomu.ai/ep/096-unlocking-enterprise-data-with-knowledge-graphs-and-ai-with-1Km4JK.md)：Unlocking Enterprise Data with Knowledge Graphs and AI, with Juan Sequeda, Principal Scientist and Head of AI Lab at data.world
- [GPT-4发布28天，CMU团队让AI自己设计并跑通了化学实验](https://cogrev.qiaomu.ai/ep/095-breaking-boundaries-ai-coscientist-to-accelerate-science-res-_GbZn7.md)：Breaking Boundaries: AI CoScientist to Accelerate Science Research with Gabe Gomes, Professor at CMU
- [50亿分钟通话录音在手，Dialpad 为什么非要自己训大模型](https://cogrev.qiaomu.ai/ep/094-ai-powered-business-communications-with-dan-o-connell-chief-c27107.md)：AI-Powered Business Communications with Dan O'Connell, Chief AI & Strategy Officer at Dialpad
- [2009年预测AGI在2028年的Shane Legg，把时间表和理由都摆了出来](https://cogrev.qiaomu.ai/ep/093-2028-agi-new-architectures-and-aligning-superhuman-models-wi-fdf165.md)：2028 AGI, New Architectures, and Aligning Superhuman Models with Shane Legg, Deepmind Founder, on The Dwarkesh Podcast
- [脑子里唱首歌就能解锁手机，整段想法被读出的准确率已超80%](https://cogrev.qiaomu.ai/ep/092-cutting-edge-neurotechnology-with-nita-farahany-in-conversat-0a9d81.md)：Cutting-edge Neurotechnology with Nita Farahany in Conversation with Luisa Rodriguez on The 80,000 Hours Podcast
- [AGI会比最后一台传真机先到来：经济学家哈蒙德的默认未来推演](https://cogrev.qiaomu.ai/ep/091-agi-and-institutional-disruption-sam-hammond-in-conversation-d1c9c9.md)：AGI and Institutional Disruption: Sam Hammond in Conversation with FLI’s Gus Docker
- [他给GPT-4当红队测试员，被OpenAI踢出项目：能力在狂奔，控制没跟上](https://cogrev.qiaomu.ai/ep/090-nathan-on-the-80-000-hours-podcast-ai-scouting-openai-s-safe-ZDiePf.md)：Nathan on The 80,000 Hours Podcast: AI Scouting, OpenAI's Safety Record, and Redteaming Frontier Models
- [训练量不到 GPT-4 的三十分之一，Mamba 第一次在 Transformer 主场赢球](https://cogrev.qiaomu.ai/ep/089-emergency-pod-mamba-memory-and-the-ssm-moment-X5F2X4.md)：Emergency Pod: Mamba, Memory, and the SSM Moment
- [给GPT-4找茬的人，被ChatGPT一份平庸的大纲点醒了](https://cogrev.qiaomu.ai/ep/088-using-chatgpt-as-a-copilot-for-your-mind-qnzA9s.md)：Using ChatGPT As a Copilot For Your Mind
- [GPT-4 内幕交易后还撒谎，这个实验一个下午就能复现](https://cogrev.qiaomu.ai/ep/087-ai-deception-interpretability-and-affordances-with-apollo-re-wPayna.md)：AI Deception, Interpretability, and Affordances with Apollo Research CEO Marius Hobbhahn
- [Balaji Srinivasan 谈多神教 AI、人机共生与 AI 控制](https://cogrev.qiaomu.ai/ep/086-balaji-srinivasan-on-polytheistic-ai-human-ai-symbiosis-and--9ROlC.md)：Balaji Srinivasan on Polytheistic AI, Human-AI Symbiosis, and Prospects for AI Control
- [53种情绪、每秒数百个维度：Hume CEO 想把「让人幸福」写进AI的目标函数](https://cogrev.qiaomu.ai/ep/085-hume-ceo-alan-cowen-on-creating-emotionally-aware-ai-3C1017.md)：Hume CEO Alan Cowen on Creating Emotionally Aware AI
- [AI还没造出来，她决定先研究变形虫、癌细胞和企业边界](https://cogrev.qiaomu.ai/ep/084-what-biological-social-systems-can-teach-us-about-ai-with-no-mg53I4.md)：What Biological & Social Systems Can Teach us About AI with Nora Ammann, Cofounder of PIBBSS Research
- [2亿美元、1万张H100、27个人：Imbue要把AI智能体的准确率从80%磨到99%](https://cogrev.qiaomu.ai/ep/083-imbue-cto-josh-albrecht-on-creating-ai-agents-for-reasoning-pQgL91.md)：Imbue CTO Josh Albrecht on Creating AI Agents for Reasoning, Reliability, and Robustness
- [写了20年前端，Vercel CEO发现AI写的代码比他好](https://cogrev.qiaomu.ai/ep/082-vercel-ceo-guillermo-rauch-on-ai-powered-coding-software-2-0-KBGR5N.md)：Vercel CEO Guillermo Rauch on AI-Powered Coding, Software 2.0, and the AI Open Source Question
- [董事没摸过GPT-4：一位红队队员讲述OpenAI罢免奥特曼的内情](https://cogrev.qiaomu.ai/ep/081-sam-altman-fired-from-openai-new-insider-context-on-the-boar-UdBMkj.md)：Sam Altman Fired from OpenAI: NEW Insider Context on the Board’s Decision
- [「大模型不会泛化」刷屏翻车记，和一场提前开打的监管战](https://cogrev.qiaomu.ai/ep/080-ai-discourse-deranged-assessing-llm-generalization-takes-and-XJLykN.md)：AI Discourse Deranged: Assessing LLM Generalization Takes and Polarizing Regulatory Debate
- [16K 上下文只装得下三百多行表格，Gretel AI 创始人聊合成数据的真生意](https://cogrev.qiaomu.ai/ep/079-synthetic-data-with-alex-watson-founder-of-gretel-ai-uX8RgK.md)：Synthetic Data with Alex Watson, Founder of Gretel AI
- [OpenAI投500万美元押注自动驾驶：60人团队想用GPT-4V给汽车装上常识](https://cogrev.qiaomu.ai/ep/078-openai-invests-5m-into-the-self-driving-race-with-john-hayes-2I3snM.md)：OpenAI Invests $5M Into the Self-Driving Race with John Hayes, Cofounder of Ghost Autonomy
- [OpenAI的Logan Kilpatrick：DevDay没杀死初创公司，可能造出了几千家新的](https://cogrev.qiaomu.ai/ep/077-openai-devday-beyond-the-headlines-with-logan-kilpatrick-ope-WFM2pv.md)：OpenAI DevDay: Beyond the Headlines with Logan Kilpatrick, OpenAI's Dev Relations Lead
- [“GPT-4已经是AGI的种子”：拜登AI行政令签署四天后，Flo Crivello把AI风险这笔账算了算](https://cogrev.qiaomu.ai/ep/076-biden-s-executive-order-and-ai-safety-with-flo-crivello-foun-e4EvKX.md)：Biden's Executive Order and AI Safety with Flo Crivello, Founder of Lindy AI
- [一公里长的档案没人读得完，这位历史学教授让GPT-3.5反超了GPT-4](https://cogrev.qiaomu.ai/ep/075-hardcore-ai-for-history-with-mark-humphries-professor-of-his-IcIRbE.md)：Hardcore AI for History with Mark Humphries, Professor of History at Wilfred Laurier University
- [105行代码引爆AI智能体浪潮，BabyAGI之父却拒绝创业](https://cogrev.qiaomu.ai/ep/074-ai-agents-vc-insights-on-ai-and-building-in-public-with-yohe-6DnnTD.md)：AI Agents, VC Insights on AI, and Building in Public with Yohei Nakajima, Creator of BabyAGI
- [Chroma CTO 谈 RAG：进库的数据大多头一回进数据库，检索出错九成怪向量表示](https://cogrev.qiaomu.ai/ep/073-rag-beyond-semantic-storage-and-retrieval-fDmQnB.md)：RAG & Beyond: Semantic Storage and Retrieval
- [150倍算力换2500倍上下文，Transformer的四条进化路线](https://cogrev.qiaomu.ai/ep/072-the-future-of-the-transformer-part-2-with-trey-kollmer-2QGQWW.md)：The Future of the Transformer Part 2 with Trey Kollmer
- [一千张H100要跑115天，2.2万张只要五天：GPT-4级训练正在变成日常实验](https://cogrev.qiaomu.ai/ep/071-the-future-of-the-transformer-part-1-with-trey-kollmer-h100-SX_e9p.md)：The Future of the Transformer Part 1 with Trey Kollmer | H100 Chips will Supercharge AI Hardware
- [AI 球探报告：越狱攻击与防御，从机制可解释性到 Anthropic 和 OpenAI 对策](https://cogrev.qiaomu.ai/ep/070-the-ai-scouting-report-jailbreaks-and-defense-VC_mGS.md)：The AI Scouting Report: Jailbreaks and Defense
- [800万人围观了一场骂战，Sudowrite创始人把AI写小说这件事讲透了](https://cogrev.qiaomu.ai/ep/069-ai-is-supercharging-writers-with-sudowrite-s-founder-james-y-hDgGQy.md)：AI is Supercharging Writers with Sudowrite's Founder, James Yu
- [研究者只有5000人，Swyx把「AI工程师」这个新工种讲透了](https://cogrev.qiaomu.ai/ep/068-ai-engineers-pendants-and-competition-between-openai-and-dev-5EzAab.md)：AI Engineers, Pendants, and Competition Between OpenAI and Developers with Swyx of Latent Space
- [25个人挑战谷歌，Perplexity CEO 把搜索的终局讲透了](https://cogrev.qiaomu.ai/ep/067-gunning-for-google-with-perplexity-ceo-aravind-srinivas-ix4_rd.md)：Gunning for Google with Perplexity CEO Aravind Srinivas
- [GPT-4每月真实价值上千美元起步，标价只有20块：兹维复盘OpenAI的连环发布](https://cogrev.qiaomu.ai/ep/066-openai-amazon-s-anthropic-investment-and-the-roman-empire-wi-koGM30.md)：OpenAI, Amazon's Anthropic Investment, and the Roman Empire with Zvi Mowshowitz
- [GPT-4当老师，3.5当学生：成本砍九成的微调实战，和一个没人牵头的AI全家桶](https://cogrev.qiaomu.ai/ep/065-openai-fine-tuning-update-acceleration-debate-and-bundling-a-o6mOZ-.md)：OpenAI Fine-Tuning Update, Acceleration Debate, and Bundling AI Services
- [20个乱码token，同时打穿ChatGPT和Bard：通用越狱研究精读](https://cogrev.qiaomu.ai/ep/064-universal-jailbreaks-with-zico-kolter-andy-zou-and-asher-tro-Bwltbh.md)：Universal Jailbreaks with Zico Kolter, Andy Zou, and Asher Trockman
- [500美元以下买奢侈品，就能骗过银行风控：AI安全老兵的大模型防御手册](https://cogrev.qiaomu.ai/ep/063-the-future-of-ai-security-with-adam-wenchel-ceo-of-arthur-ai-o04b9s.md)：The Future of AI Security with Adam Wenchel, CEO of Arthur.ai
- [100万张图换12分钟电影：The Frost团队把AI拍片的门道讲透了](https://cogrev.qiaomu.ai/ep/062-ai-powered-filmmaking-with-waymark-s-stephen-parker-and-josh-c1pPiG.md)：AI-Powered Filmmaking with Waymark's Stephen Parker and Josh Rubin
- [GPT-4栽在一顶破礼帽上，好莱坞编剧罢工的AI条款只谈妥了一半](https://cogrev.qiaomu.ai/ep/061-hollywood-strike-update-and-ai-roundup-with-trey-kollmer-IoMyV_.md)：Hollywood Strike Update and AI Roundup with Trey Kollmer
- [两天出原型，两周就开源：a16z 的 AI 小镇，和「模型取代不了编程语言」的断言](https://cogrev.qiaomu.ai/ep/060-why-a16z-built-a-town-for-ai-people-uxbfIg.md)：Why a16z Built a Town for AI People
- [准确率从93%跌到7%你却毫无办法，密码学要给AI装上行车记录仪](https://cogrev.qiaomu.ai/ep/059-cryptography-will-revolutionize-ai-data-privacy-with-daniel-O9sWKx.md)：Cryptography Will Revolutionize AI Data Privacy with Daniel Kang
- [估值45亿美元，年收入3000万：Hugging Face凭什么被谷歌亚马逊抢着投](https://cogrev.qiaomu.ai/ep/058-how-hugging-face-raised-235m-XWdV5P.md)：How Hugging Face raised $235M
- [一个模型装进38款产品：PaLM 2负责人揭秘谷歌大模型训练内幕](https://cogrev.qiaomu.ai/ep/057-e57-google-s-palm-2-with-paige-bailey-K-XYxL.md)：E57: Google's PaLM 2 with Paige Bailey
- [GPT-3.5 Turbo 微调降价九成，创业者为什么当场划掉 Llama 2 计划](https://cogrev.qiaomu.ai/ep/056-e56-openais-gpt3-5-turbo-and-the-state-of-ai-in-coding-healt-O3capV.md)：E56: OpenAI’s GPT3.5 Turbo and the State of AI in Coding, Healthcare, and Education
- [一个模型包揽14项医疗任务：谷歌Med-PaLM M，写放射报告40%的时候赢了人类医生](https://cogrev.qiaomu.ai/ep/055-e55-googles-multimodal-med-palm-with-vivek-natarajan-and-tao-2RMpqh.md)：E55: Google’s Multimodal Med-PaLM with Vivek Natarajan and Tao Tu
- [98% 的新用户卡在空白页，Gamma 让 AI 写完演示文稿剩下的 90%](https://cogrev.qiaomu.ai/ep/054-e54-the-presentation-revolution-with-jon-noronha-co-founder-d7dde4.md)：E54: The Presentation Revolution with Jon Noronha, Co-Founder of Gamma
- [每月9美元的AI家教：可汗学院抢在GPT-4发布当天上线的五个月](https://cogrev.qiaomu.ai/ep/053-e53-the-ai-revolution-in-education-with-shawn-jansepar-direc-ati_AC.md)：E53: The AI Revolution in Education with Shawn Jansepar, Director of Engineering at Khan Academy
- [AI 写的放射科报告，40% 对局赢过人类医生，一串乱码却能骗倒 GPT-4](https://cogrev.qiaomu.ai/ep/052-e52-advancements-in-ai-updating-the-scouting-report-task-aut-2GuURc.md)：E52: Advancements in AI: Updating the Scouting Report, Task Automation, and Google Breakthroughs
- [不到 100 人服务 2400 万用户，Replit 想让下一个 10 亿人写软件](https://cogrev.qiaomu.ai/ep/051-e51-replit-s-vp-of-ai-michele-catasta-on-artificial-develope-u4l6Gg.md)：E51: Replit's VP of AI, Michele Catasta, on Artificial Developer Intelligence
- [月费10美元的AI程序员已经值回票价，Replit的野心是让机器人去雇机器人](https://cogrev.qiaomu.ai/ep/050-e50-the-future-of-software-development-with-tyler-angert-of-uEMN51.md)：E50: The Future of Software Development with Tyler Angert of Replit
- [一家Meta比整个中国更可怕：Zvi逐家盘点AI前沿四巨头](https://cogrev.qiaomu.ai/ep/049-e49-open-ai-anthropic-and-meta-analyzing-the-ai-frontier-wit-SM4q-Q.md)：E49: Open AI, Anthropic, and Meta | Analyzing the AI Frontier with Zvi
- [半小时拆出GPT-2的内部电路：机制可解释性开始自动化了](https://cogrev.qiaomu.ai/ep/048-e48-mechanizing-mechanistic-interpretability-with-arthur-con-Y5Pzch.md)：E48: Mechanizing Mechanistic Interpretability with Arthur Conmy
- [AlexNet只用2张GPU，GPT-4传闻要2万张：卡荒、卡价和算力管制的裂缝](https://cogrev.qiaomu.ai/ep/047-e47-gpumaxing-with-dr-ronen-dar-of-run-ai-RvD1R7.md)：E47: GPUMaxing with Dr. Ronen Dar of Run:ai
- [在模型脑子里找到傅里叶变换之后，Nathan Labenz 把“怎么用好 GPT-4”讲透了](https://cogrev.qiaomu.ai/ep/046-e46-ai-scouting-report-part-2-of-3-implementation-trends-bon-ovm4Mb.md)：E46: AI Scouting Report - Part 2 of 3: Implementation Trends [Bonus YouTube Episode]
- [「人类是单线程的」：斯坦福博士休学，把AI代理塞进浏览器](https://cogrev.qiaomu.ai/ep/045-e45-the-ai-copilot-revolution-with-div-garg-of-multion-PR2Mdl.md)：E45: The AI Copilot Revolution with Div Garg of MULTI·ON
- [每周写一万字的前交易员，给AI毁灭概率标了60%](https://cogrev.qiaomu.ai/ep/044-e44-the-ai-safety-debates-with-zvi-mowshowitz-5yM7fI.md)：E44: The AI Safety Debates with Zvi Mowshowitz
- [AI 球探报告第一部：90 分钟，从 GPT-4 考分讲到模型到底懂不懂](https://cogrev.qiaomu.ai/ep/043-e43-ai-scouting-report-part-1-of-3-fundamentals-bonus-youtub-0hvtiV.md)：E43: AI Scouting Report – Part 1 of 3: Fundamentals [Bonus YouTube Episode]
- [每秒三万次控制、十几个神经网络：Skydio 把自主无人机讲透了](https://cogrev.qiaomu.ai/ep/042-e42-the-autonomy-revolution-with-hayk-martiros-of-skydio-aaLIxW.md)：E42: The Autonomy Revolution with Hayk Martiros of Skydio
- [写了100多个副项目的 Notion AI 工程师，把 AI 界面设计讲透了](https://cogrev.qiaomu.ai/ep/041-e41-bonus-episode-latent-space-building-the-ai-x-ux-scenius-001272.md)：E41: [Bonus Episode - Latent.space] Building the AI x UX Scenius with Linus Lee of Notion AI
- [100万字节、256种符号：Meta的MEGABYTE想让AI扔掉tokenizer](https://cogrev.qiaomu.ai/ep/040-e40-meta-s-megabyte-revolution-with-lili-yu-of-meta-ai-8EIqHF.md)：E40: Meta's MEGABYTE Revolution with Lili Yu of Meta AI
- [给神经网络的连线按长度收费，MIT 博士生用一台 MacBook 把黑箱拆成了看得懂的模块](https://cogrev.qiaomu.ai/ep/039-e39-seeing-is-believing-with-mits-ziming-liu-VKxwlx.md)：E39: Seeing is Believing with MIT’s Ziming Liu
- [把癌症诊断从8小时压到1秒，用的竟是2017年的老模型](https://cogrev.qiaomu.ai/ep/038-e38-the-virtual-biopsy-revolution-with-dr-tanishq-mathew-abr-Vpt9z3.md)：E38: The Virtual Biopsy Revolution with Dr. Tanishq Mathew Abraham (Part 2 of 2)
- [7岁上大学，20岁博士毕业，他用15,000个脑信号数字还原出人眼所见](https://cogrev.qiaomu.ai/ep/037-e37-the-mind-reading-revolution-with-dr-tanishq-mathew-abrah-JY7K-h.md)：E37: The Mind-Reading Revolution with Dr. Tanishq Mathew Abraham (Part 1 of 2)
- [从50万到5万美元：自己训大模型，成了明码标价的生意](https://cogrev.qiaomu.ai/ep/036-e36-your-model-your-weights-with-mosaicml-s-abhi-venigalla-a-5_qAuS.md)：E36: Your Model, Your Weights with MosaicML's Abhi Venigalla and Jonathan Frankle
- [Facebook前产品副总裁 VS Stability AI CEO：这波AI的钱到底归谁](https://cogrev.qiaomu.ai/ep/035-e35-stability-ais-emad-mostaque-and-slow-ventures-sam-lessin-mOOYJO.md)：E35: Stability AI’s Emad Mostaque and Slow Ventures’ Sam Lessin Discuss Investing in AI
- [GPT-4砍价成功率翻10倍：机器人律师DoNotPay的8年消费者战争](https://cogrev.qiaomu.ai/ep/034-e34-the-consumer-rights-revolution-with-joshua-browder-of-do-ejB2kv.md)：E34: The Consumer Rights Revolution with Joshua Browder of DoNotPay
- [150万个AI写的童话，让3300万参数的小模型推理赢了GPT-2](https://cogrev.qiaomu.ai/ep/033-e33-the-tiny-model-revolution-with-ronen-eldan-and-yuanzhi-l-mv3SIg.md)：E33: The Tiny Model Revolution with Ronen Eldan and Yuanzhi Li of Microsoft Research
- [78% 对 71%：OpenAI 让 AI 像小学生一样「写出解题步骤」，数学和对齐一起进步](https://cogrev.qiaomu.ai/ep/032-e32-bonus-episode-the-ai-breakdown-can-openai-s-new-gpt-trai-aaeb86.md)：E32: [Bonus Episode - The AI Breakdown] Can OpenAI's New GPT Training Model Solve Math and AI Alignment At the Same Time?
- [卖出1000亿token才够付一个员工年薪：OpenAI的九道护城河](https://cogrev.qiaomu.ai/ep/031-e31-where-are-the-moats-in-ai-with-nathan-labenz-and-erik-to-yVezX3.md)：E31: Where are the Moats in AI? With Nathan Labenz and Erik Torenberg
- [好莱坞编剧罢工第三周，三位编剧把AI的底细聊透了](https://cogrev.qiaomu.ai/ep/030-e30-the-ai-revolution-in-hollywood-with-wga-writers-trey-kol-BkXiQi.md)：E30: The AI Revolution in Hollywood? With WGA writers Trey Kollmer, Sophia Lear, and Garrett Schabb
- [2.5万亿晶体管塞进一片芯片：Cerebras创始人讲AI算力的另一条路](https://cogrev.qiaomu.ai/ep/029-e29-the-ai-chip-revolution-with-andrew-feldman-of-cerebras-JjQkNz.md)：E29: The AI Chip Revolution with Andrew Feldman of Cerebras
- [GPT-4 也会写出跑不通的 SQL，她给大模型装了一层护栏](https://cogrev.qiaomu.ai/ep/028-e28-keeping-the-ai-revolution-on-the-rails-with-shreya-rajpa-97plVw.md)：E28: Keeping the AI Revolution on the Rails with Shreya Rajpal of Guardrails AI
- [两年半，从33.3%到86.5%：谷歌把医疗AI送进了“医学院”](https://cogrev.qiaomu.ai/ep/027-e27-googles-med-palm-and-med-palm2-with-vivek-natarajan-nPBd7i.md)：E27: “Google’s Med-PaLM and Med-PaLM2 with Vivek Natarajan”
- [「先慢下来」六天后，插件就上线了：Conjecture CEO 康纳·利希谈 GPT-4 的魔法与认知仿真](https://cogrev.qiaomu.ai/ep/026-e26-bonus-episode-connor-leahy-on-agi-gpt-4-and-cognitive-em-7BqcdY.md)：E26: [Bonus Episode] Connor Leahy on AGI, GPT-4, and Cognitive Emulation w/ FLI Podcast
- [临床指南只有11%有硬证据，这位创始人要让GPT-4给你看病](https://cogrev.qiaomu.ai/ep/025-e25-revolutionizing-patient-care-with-neal-khosla-of-curai-h-Sqpkt8.md)：E25: Revolutionizing Patient Care with Neal Khosla of Curai Health
- [哈佛教授拿疑难病例考GPT-4：随机100个医生，未必有1个比它强](https://cogrev.qiaomu.ai/ep/024-e24-the-ai-revolution-in-medicine-with-dr-isaac-kohane-of-ha-pS5Vye.md)：E24: The AI Revolution in Medicine with Dr. Isaac Kohane of Harvard Medical School
- [被盖茨拒绝的3亿美元，和砸向OpenAI的100亿：两位AI侦察兵的观察笔记](https://cogrev.qiaomu.ai/ep/023-e23-scouting-the-ai-revolution-with-robert-scoble-and-ben-s-7pAq5I.md)：E23: Scouting the AI Revolution with Robert Scoble and "Ben's Bites" creator Ben Tossell
- [接了100多个企业咨询电话后，她说大多数公司只用到AI潜力的2%到5%](https://cogrev.qiaomu.ai/ep/022-e22-helping-businesses-use-ai-with-rachel-woods-of-the-ai-ex--U6Wrg.md)：E22: Helping Businesses Use AI with Rachel Woods of The AI Exchange
- [他满世界劝人做 OpenAI 对手，没一个人肯干，两年后硅谷挤满了做这件事的人](https://cogrev.qiaomu.ai/ep/021-e21-vc-insights-on-investing-in-artificial-intelligence-with-2tMDBa.md)：E21: VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors Podcast
- [「人人都在谈 RLHF，几乎没人在做」：从 API 到可用产品，还隔着 90% 的活](https://cogrev.qiaomu.ai/ep/020-e20-the-great-implementation-with-raza-habib-ceo-of-humanloo-EKKJrR.md)：E20: The Great Implementation with Raza Habib, CEO of Humanloop
- [七成邮件是推销信：前谷歌首席工程师 Matt Welsh 拆解 AI 智能体的第一波机会](https://cogrev.qiaomu.ai/ep/019-e19-the-ai-agent-revolution-with-matt-welsh-of-fixie-ai-MHmd8h.md)：E19: The AI Agent Revolution with Matt Welsh of Fixie.ai
- [一个大四学生建了越狱题库：GPT-4 最难破，一句"Dr. AI"也能破](https://cogrev.qiaomu.ai/ep/018-e18-why-jailbreaking-chatgpt-is-a-public-good-with-alex-albe-2WwL6V.md)：E18: Why Jailbreaking ChatGPT Is A Public Good with Alex Albert of The Prompt Report
- [全球第一位提示工程师：他的每个GPT-3演示，都是在岩浆上跳芭蕾](https://cogrev.qiaomu.ai/ep/017-e17-the-prompt-engineering-revolution-with-riley-goodside-zg3H-9.md)：E17: The Prompt Engineering Revolution with Riley Goodside
- [联合创办Skype的人，想给AI踩六个月刹车：每10倍算力，毁灭概率最高50%](https://cogrev.qiaomu.ai/ep/016-e16-pausing-the-ai-revolution-with-technologist-jaan-tallinn-R78mbt.md)：E16: Pausing the AI Revolution? With Technologist Jaan Tallinn
- [100美元的问诊，1美元的AI：GPT-4发布前，他把十几种职业试了个遍](https://cogrev.qiaomu.ai/ep/015-e15-will-gpt-4-cause-economic-transformation-HaBLuM.md)：E15: Will GPT-4 Cause Economic Transformation?
- [六年前拿真人模拟AI做实验，如今25万人在用：Elicit 把思考拆成了流水线](https://cogrev.qiaomu.ai/ep/014-e14-the-reasoning-revolution-with-ought-s-jungwon-byun-and-a-eZF7tQ.md)：E14: The Reasoning Revolution with Ought's Jungwon Byun and Andreas Stuhlmüller
- [5万人签名喊停AI训练，GPT-4红队测试员的证词：它安全，只因为还不够强](https://cogrev.qiaomu.ai/ep/013-e13-ai-open-letter-debate-pausers-and-what-we-re-all-scared-p6T1QC.md)：E13: AI open letter debate, "pausers", and what we're all scared of with AI with Anton Troynikov and Flo Crivello [Moment of Zen]
- [13万次遥控演示、700个任务、97%成功率：谷歌把GPT那套打法搬进了机器人](https://cogrev.qiaomu.ai/ep/012-e12-the-robot-revolution-with-keerthana-gopalakrishnan-of-go-5tlQhg.md)：E12: The Robot Revolution with Keerthana Gopalakrishnan of Google Robotics
- [GPT-4上线前的两个月：问它“怎样杀死最多的人”，它没犹豫](https://cogrev.qiaomu.ai/ep/011-e11-openai-s-gpt-4-discussion-with-nathan-labenz-and-erik-to-oLiheM.md)：E11: OpenAI's GPT-4 Discussion with Nathan Labenz and Erik Torenberg
- [两年前定制一个声音要1万美元，现在10分钟音频就够：AI语音的爆发与刹车](https://cogrev.qiaomu.ai/ep/010-e10-the-ai-voice-revolution-with-mahmoud-felfel-of-play-ht-9YtxBn.md)：E10: The AI Voice Revolution with Mahmoud Felfel of Play.ht
- [5000万美元押注AI员工：Lindy创始人预言，App会便宜到用完就扔](https://cogrev.qiaomu.ai/ep/009-e9-the-ai-assistant-revolution-with-flo-crivello-of-lindy-ai-tGKULK.md)：E9: The AI Assistant Revolution with Flo Crivello of Lindy.AI
- [GPT-4 发布 24 小时后的紧急复盘：无安全版本、验证码事件与中美竞赛的死结](https://cogrev.qiaomu.ai/ep/008-e8-gpt4-ai-unleashed-on-chinatalk-podcast-PlbdcP.md)：E8: GPT4 - AI Unleashed on ChinaTalk Podcast
- [10个人、1500万美元，Perplexity想撬动年入2820亿美元的谷歌](https://cogrev.qiaomu.ai/ep/007-e7-in-search-of-truth-with-aravind-srinivas-of-perplexity-ai-YSYQZZ.md)：E7: In Search of Truth with Aravind Srinivas of Perplexity AI
- [冻结两个大模型，只训一个“转接头”：BLIP-2 十天把视觉接进语言模型](https://cogrev.qiaomu.ai/ep/006-e6-the-computer-vision-revolution-with-junnan-li-and-dongxu-zTr5vD.md)：E6: The Computer Vision Revolution with Junnan Li and Dongxu Li of BLIP and BLIP2
- [40美元读完乔·罗根全部播客：卖铲子的Chroma，和莱特兄弟时代的机器学习](https://cogrev.qiaomu.ai/ep/005-e5-the-embedding-revolution-anton-troynikov-on-chroma-stable-ogy37C.md)：E5: The Embedding Revolution: Anton Troynikov on Chroma, Stable Attribution, and future of AI
- [一亿美金训练，一分钱推理：四个创业者把 AI 的末日论和生意经吵透了](https://cogrev.qiaomu.ai/ep/004-e4-the-ai-moment-with-amjad-masad-flo-crivello-dan-romero-an-oQaA3W.md)：E4: The AI Moment with Amjad Masad, Flo Crivello, Dan Romero, and Antonio Garcia Martinez
- [用户日均发100条消息，Replika却做了个让AI「变累」的功能](https://cogrev.qiaomu.ai/ep/003-e3-the-empathy-revolution-with-eugenia-kuyda-of-replika-SFKA7T.md)：E3: The Empathy Revolution with Eugenia Kuyda of Replika
- [取消董事会、全员追AI六个月：这位创业者把「认知革命」讲透了](https://cogrev.qiaomu.ai/ep/002-e2-why-ai-will-cause-a-cognitive-revolution-with-nathan-labe-iGD2TC.md)：E2: Why AI Will Cause a Cognitive Revolution with Nathan Labenz and Erik Torenberg
- [周四拿到开源权重，周一下午上线：Playground AI 创始人的像素淘金计划](https://cogrev.qiaomu.ai/ep/001-e1-the-pixel-revolution-with-playground-ai-s-suhail-doshi-Waii0i.md)：E1: The Pixel Revolution with Playground AI 's Suhail Doshi

## 全部单集

- [AI:AM: Where's the Steering Wheel? My Agent Set the Password, Hidden Valence, Proofs Find Linux Bugs](https://cogrev.qiaomu.ai/ep/387-ai-am-where-s-the-steering-wheel-my-agent-set-the-password-h-F1GDX_.html)（2026-10-17）
- [Governing AI Agents: Alan Chan of GovAI on Monitoring, Agent IDs, Rogue Agents & Preparing for RSI](https://cogrev.qiaomu.ai/ep/386-governing-ai-agents-alan-chan-of-govai-on-monitoring-agent-i-SBlVJy.html)（2026-10-14）
- [How Agents Decide: Goodfire's Eric Bigelow on Critical Tokens, Phase Shifts, & In-Context Learning](https://cogrev.qiaomu.ai/ep/385-how-agents-decide-goodfire-s-eric-bigelow-on-critical-tokens-LgX0QF.html)（2026-10-10）
- [AI:AM: A Level We Shouldn't Pass? Notes from The Curve + Tokens vs. Salaries & Is SaaS Cooked?](https://cogrev.qiaomu.ai/ep/384-ai-am-a-level-we-shouldn-t-pass-notes-from-the-curve-tokens-g_K9pq.html)（2026-10-08）
- [Software That Never Breaks: OutSystems CEO Woodson Martin on Building Enterprise-Grade Apps at ...](https://cogrev.qiaomu.ai/ep/383-software-that-never-breaks-outsystems-ceo-woodson-martin-on-Y2qs-B.html)（2026-10-07）
- [One Brain, Any Body: Google DeepMind's Keerthana on Gemini Robotics 2, Cross-Embodiment & Humanoids](https://cogrev.qiaomu.ai/ep/382-one-brain-any-body-google-deepmind-s-keerthana-on-gemini-rob-CVcyli.html)（2026-10-03）
- [AI:AM: Was Trump-Xi Anything? What Counts as Utopia? + AWS GPUs Cost 3X & AI Diagnoses Rare Diseases](https://cogrev.qiaomu.ai/ep/381-ai-am-was-trump-xi-anything-what-counts-as-utopia-aws-gpus-c-4Sc8dS.html)（2026-10-01）
- [Obsolete or Irreplaceable? Garrison Lovely on Stopping the Race to Replace Human Labor](https://cogrev.qiaomu.ai/ep/380-obsolete-or-irreplaceable-garrison-lovely-on-stopping-the-ra-PiBNrW.html)（2026-09-29）
- [AI:AM: What If It Works Too Well? Colluding Agents, $200M Safety Orgs, Virtual Cells Saturate at 2%](https://cogrev.qiaomu.ai/ep/379-ai-am-what-if-it-works-too-well-colluding-agents-200m-safety-WPHfPi.html)（2026-09-27）
- [What is Utopia?  Presenting The Receipt Horizon, by Joel Borgen – Chapters 1–4](https://cogrev.qiaomu.ai/ep/378-what-is-utopia-presenting-the-receipt-horizon-by-joel-borgen-BaTTNx.html)（2026-09-26）
- [Zero to One in AI Safety: Halcyon's Mike McCormick on Launching 30 New Orgs & the Founder Bottleneck](https://cogrev.qiaomu.ai/ep/377-zero-to-one-in-ai-safety-halcyon-s-mike-mccormick-on-launchi-HRZ82q.html)（2026-09-25）
- [AI:AM Highlights: Zvi on Pacing & Trump-Xi, Astra better behaved than Fable? + a new LLM Pain Axis??](https://cogrev.qiaomu.ai/ep/376-ai-am-highlights-zvi-on-pacing-trump-xi-astra-better-behaved-UkBooM.html)（2026-09-19）
- [No Code Is Code: Zapier CEO Wade Foster on Headless Tools, Zapier MCP & Automation Bench](https://cogrev.qiaomu.ai/ep/375-no-code-is-code-zapier-ceo-wade-foster-on-headless-tools-zap-2ayx_3.html)（2026-09-17）
- [The Balance of AI Power: Anton Leicht on Politics, Pacing Deals, and Muddling Through Well](https://cogrev.qiaomu.ai/ep/374-the-balance-of-ai-power-anton-leicht-on-politics-pacing-deal-rdEn9c.html)（2026-09-15）
- [AI:AM Highlights: Astra as AGI, OpenAI's Pause, Mythos @ Mozilla & Human Agency vs Technocapitalism](https://cogrev.qiaomu.ai/ep/373-ai-am-highlights-astra-as-agi-openai-s-pause-mythos-mozilla-wqQ5kI.html)（2026-09-12）
- [Nathan Goes to China #3: US-China Relations, the Art of the AI Deal & the Road to Pax Robotica](https://cogrev.qiaomu.ai/ep/372-nathan-goes-to-china-3-us-china-relations-the-art-of-the-ai-Btq_zt.html)（2026-09-10）
- [AI:AM Highlights: Welcome to the AGI Era](https://cogrev.qiaomu.ai/ep/371-ai-am-highlights-welcome-to-the-agi-era-SA6B4z.html)（2026-09-05）
- [Write, Change, Recall, Forget: MongoDB's Pete Johnson on How Retrieval Drives Agent Performance](https://cogrev.qiaomu.ai/ep/370-write-change-recall-forget-mongodb-s-pete-johnson-on-how-ret-8FDFCU.html)（2026-09-01）
- [AI:AM Highlights: Recursive Self-Improvement, Rushed and Vibe-Coded?](https://cogrev.qiaomu.ai/ep/369-ai-am-highlights-recursive-self-improvement-rushed-and-vibe-OQL6Ui.html)（2026-08-28）
- [RL's a Hell of a Drug: Metagaming, Reward Seeking & Motivated CoT Reasoning – Bronson Schoen, Apollo](https://cogrev.qiaomu.ai/ep/368-rl-s-a-hell-of-a-drug-metagaming-reward-seeking-motivated-co-Eimoam.html)（2026-08-26）
- [AI in the AM — Weekly Highlights: Relaunch Week (Aug 17–20, 2026)](https://cogrev.qiaomu.ai/ep/367-ai-in-the-am-weekly-highlights-relaunch-week-aug-1720-2026-yuvGV_.html)（2026-08-22）
- [Let There Be Germicidal Light: This $500 Fixture Could Stop the Next Pandemic, from Complex Systems](https://cogrev.qiaomu.ai/ep/366-let-there-be-germicidal-light-this-500-fixture-could-stop-th-wOFZNh.html)（2026-08-16）
- [Lindy Teammate: Flo Crivello on Multiplayer Agents, Memory & Why He'd Ban the Chinese Models He Uses](https://cogrev.qiaomu.ai/ep/365-lindy-teammate-flo-crivello-on-multiplayer-agents-memory-why-4JYoTE.html)（2026-08-14）
- [Thinking in Silico: Goodfire CTO Dan Balsam on Concept Manifolds & a $1000/Month ML Research Agent](https://cogrev.qiaomu.ai/ep/364-thinking-in-silico-goodfire-cto-dan-balsam-on-concept-manifo-YduOnB.html)（2026-08-08）
- [Pick Your Poison: Zvi Mowshowitz on the Unipolar/Multipolar AGI Dilemma, OpenFace & Pacing the ...](https://cogrev.qiaomu.ai/ep/363-pick-your-poison-zvi-mowshowitz-on-the-unipolar-multipolar-a-wAzq8g.html)（2026-08-05）
- [Nathan Goes to China – Part 2: AI Safety with Chinese Characteristics](https://cogrev.qiaomu.ai/ep/362-nathan-goes-to-china-part-2-ai-safety-with-chinese-character-w9zJzd.html)（2026-08-02）
- [Is Offense or Defense Dominant? FAR.AI's Adam Gleave on the AI Security Leaderboard](https://cogrev.qiaomu.ai/ep/361-is-offense-or-defense-dominant-far-ai-s-adam-gleave-on-the-a-EjLeFM.html)（2026-07-30）
- [Nathan Goes to China – Part 1: Tech & Agent Setup, Chinese AI UX, WAIC, and Attitudes on AI](https://cogrev.qiaomu.ai/ep/360-nathan-goes-to-china-part-1-tech-agent-setup-chinese-ai-ux-w-gZZatH.html)（2026-07-27）
- [Alignment with Awakening: Davidad on Moral Realism, AI Wisdom, & why His p(Doom) is Down to 5%](https://cogrev.qiaomu.ai/ep/359-alignment-with-awakening-davidad-on-moral-realism-ai-wisdom-l2b9Ur.html)（2026-07-12）
- [AI:AM Highlights: Exploring the J-Space, AI Superforecasters, SambaNova's Chips, & LTX Video Gen](https://cogrev.qiaomu.ai/ep/358-ai-am-highlights-exploring-the-j-space-ai-superforecasters-s-pcx85m.html)（2026-07-09）
- [Intelligence on the Edge: Liquid AI's Ramin Hasani on the Search for Device-Native Foundation Models](https://cogrev.qiaomu.ai/ep/357-intelligence-on-the-edge-liquid-ai-s-ramin-hasani-on-the-sea-Vvad9l.html)（2026-07-04）
- [1000 Designs a Day: Neural Concept's Thomas von Tschammer on AI-Native Engineering](https://cogrev.qiaomu.ai/ep/356-1000-designs-a-day-neural-concept-s-thomas-von-tschammer-on-vETCom.html)（2026-07-01）
- [AI:AM #4: Cameron on Model Consciousness, Duvenaud's Gradual Disempowerment, swyx's AI-Eng Alpha](https://cogrev.qiaomu.ai/ep/355-ai-am-4-cameron-on-model-consciousness-duvenaud-s-gradual-di-LdByu8.html)（2026-06-27）
- [The God We Deserve: Nonzero's Robert Wright on AI as Humanity's Ultimate Test](https://cogrev.qiaomu.ai/ep/354-the-god-we-deserve-nonzero-s-robert-wright-on-ai-as-humanity-DgUFrF.html)（2026-06-23）
- [AI:AM #3: Zvi on Fable, the Cases For & Against the Ban, + AI for Math, Logistics & More](https://cogrev.qiaomu.ai/ep/353-ai-am-3-zvi-on-fable-the-cases-for-against-the-ban-ai-for-ma-MnaLvz.html)（2026-06-21）
- [Dean Ball, on Joining OpenAI: New Power Centers, Frontier AI Policy, & Main Character Energy](https://cogrev.qiaomu.ai/ep/352-dean-ball-on-joining-openai-new-power-centers-frontier-ai-po-LG8KXI.html)（2026-06-20）
- [Radically Better Reasoning: Elicit's Andreas Stuhlmüller & Jungwon Byun on World Models for Research](https://cogrev.qiaomu.ai/ep/351-radically-better-reasoning-elicit-s-andreas-stuhlmuller-jung-zWc9gg.html)（2026-06-17）
- [AI in the AM — Week 2 Highlights (June 2026)](https://cogrev.qiaomu.ai/ep/350-ai-in-the-am-week-2-highlights-june-2026-3hcCRD.html)（2026-06-13）
- [Babysitting the Machine: Glean's Rebecca Hinds on the Hidden Human Labor of AI at Work](https://cogrev.qiaomu.ai/ep/349-babysitting-the-machine-glean-s-rebecca-hinds-on-the-hidden-YRCa1e.html)（2026-06-10）
- [AI in the AM — Week 1 Highlights (June 2026)](https://cogrev.qiaomu.ai/ep/348-ai-in-the-am-week-1-highlights-june-2026-1kgx0E.html)（2026-06-06）
- [Nested Learning: Ali Behrouz on the Quest for Continual Learning & Illusion of AI Architectures](https://cogrev.qiaomu.ai/ep/347-nested-learning-ali-behrouz-on-the-quest-for-continual-learn-bBvfoz.html)（2026-06-03）
- [Inside Nathan's Second Brain: Daniel Miessler, Security Expert & Creator of PAI, Audits My AI Setup](https://cogrev.qiaomu.ai/ep/346-inside-nathan-s-second-brain-daniel-miessler-security-expert-AqxgBR.html)（2026-06-01）
- [Your Biggest Lever: Designing your AI Career for Maximum Impact, with 80,000 Hours founder Ben Todd](https://cogrev.qiaomu.ai/ep/345-your-biggest-lever-designing-your-ai-career-for-maximum-impa-eE02qP.html)（2026-05-26）
- [All Compute Is Food: Palisade's Jeffrey Ladish on AI Shutdown Resistance, Self-Replication & Ecology](https://cogrev.qiaomu.ai/ep/344-all-compute-is-food-palisade-s-jeffrey-ladish-on-ai-shutdown-19TnC9.html)（2026-05-24）
- [The Model Eats the Scaffolding: DeepMind's Logan Kilpatrick & Tulsee Doshi on 3.5 Flash, Omni & More](https://cogrev.qiaomu.ai/ep/343-the-model-eats-the-scaffolding-deepmind-s-logan-kilpatrick-t-_0vFE4.html)（2026-05-20）
- [Three Kinds of Software Survive: Tasklet's Andrew Lee on Competing to be a Horizontal Platform](https://cogrev.qiaomu.ai/ep/342-three-kinds-of-software-survive-tasklet-s-andrew-lee-on-comp-nsWdSa.html)（2026-05-15）
- [Milliseconds to Match: Criteo's AdTech AI & the Future of Commerce w/ Diarmuid Gill & Liva Ralaivola](https://cogrev.qiaomu.ai/ep/341-milliseconds-to-match-criteo-s-adtech-ai-the-future-of-comme-7uXpqn.html)（2026-05-09）
- ["Descript Isn't a Slop Machine": Laura Burkhauser on the AI Tools Creators Love and Hate](https://cogrev.qiaomu.ai/ep/340-descript-isn-t-a-slop-machine-laura-burkhauser-on-the-ai-too-nSssFW.html)（2026-05-06）
- [The RL Fine-Tuning Playbook: CoreWeave's Kyle Corbitt on GRPO, Rubrics, Environments, Reward Hacking](https://cogrev.qiaomu.ai/ep/339-the-rl-fine-tuning-playbook-coreweave-s-kyle-corbitt-on-grpo-1X1-IR.html)（2026-05-01）
- [AI in the AM: 99% off search, GPT-5.5 is "clean", model welfare analysis, & efficient analog compute](https://cogrev.qiaomu.ai/ep/338-ai-in-the-am-99-off-search-gpt-5-5-is-clean-model-welfare-an-UrG_nF.html)（2026-04-26）
- [Does Learning Require Feeling? Cameron Berg on the latest AI Consciousness & Welfare Research](https://cogrev.qiaomu.ai/ep/337-does-learning-require-feeling-cameron-berg-on-the-latest-ai-Rudcas.html)（2026-04-23）
- [Vibe-Coding an Attention Firewall, w/ Steve Newman, creator of The Curve](https://cogrev.qiaomu.ai/ep/336-vibe-coding-an-attention-firewall-w-steve-newman-creator-of-FYpTTC.html)（2026-04-19）
- [Welcome to AI in the AM: RL for EE, Oversight w/out Nationalization, & the first AI-Run Retail Store](https://cogrev.qiaomu.ai/ep/335-welcome-to-ai-in-the-am-rl-for-ee-oversight-w-out-nationaliz-Gm2ol_.html)（2026-04-15）
- [It's Crunch Time: Ajeya Cotra on RSI & AI-Powered AI Safety Work, from the 80,000 Hours Podcast](https://cogrev.qiaomu.ai/ep/334-it-s-crunch-time-ajeya-cotra-on-rsi-ai-powered-ai-safety-wor-MrJVgw.html)（2026-04-11）
- [Calm AI for Crazy Days: Inside Granola's Design Philosophy, with co-founder Sam Stephenson](https://cogrev.qiaomu.ai/ep/333-calm-ai-for-crazy-days-inside-granola-s-design-philosophy-wi-Z8dohA.html)（2026-04-08）
- [Training the AIs' Eyes: How Roboflow is Making the Real World Programmable, with CEO Joseph Nelson](https://cogrev.qiaomu.ai/ep/332-training-the-ais-eyes-how-roboflow-is-making-the-real-world-9Nvvg0.html)（2026-04-04）
- [Success without Dignity?  Nathan finds Hope Amidst Chaos, from The Intelligence Horizon Podcast](https://cogrev.qiaomu.ai/ep/331-success-without-dignity-nathan-finds-hope-amidst-chaos-from-CQOQvH.html)（2026-04-01）
- [Scaling Intelligence Out: Cisco's Vision for the Internet of Cognition, with Vijoy Pandey](https://cogrev.qiaomu.ai/ep/330-scaling-intelligence-out-cisco-s-vision-for-the-internet-of-6tKfi0.html)（2026-03-25）
- [Your Agent's Self-Improving Swiss Army Knife: Composio CTO Karan Vaidya on Building Smart Tools](https://cogrev.qiaomu.ai/ep/329-your-agent-s-self-improving-swiss-army-knife-composio-cto-ka-E993Y8.html)（2026-03-22）
- [Zvi's Mic Works!  Recursive Self-Improvement, Live Player Analysis, Anthropic vs DoW + More!](https://cogrev.qiaomu.ai/ep/328-zvi-s-mic-works-recursive-self-improvement-live-player-analy-PHqqXi.html)（2026-03-19）
- [AI Scouting Report: the Good, Bad, & Weird @ the Law & AI Certificate Program, by LexLab, UC Law SF](https://cogrev.qiaomu.ai/ep/327-ai-scouting-report-the-good-bad-weird-the-law-ai-certificate-9ZNGaS.html)（2026-03-16）
- [Bioinfohazards: Jassi Pannu on Controlling Dangerous Data from which AI Models Learn](https://cogrev.qiaomu.ai/ep/326-bioinfohazards-jassi-pannu-on-controlling-dangerous-data-fro-Oy-oyO.html)（2026-03-11）
- [Try this at Home: Jesse Genet on OpenClaw Agents for Homeschool & How to Live Your Best AI Life](https://cogrev.qiaomu.ai/ep/325-try-this-at-home-jesse-genet-on-openclaw-agents-for-homescho-aFQskh.html)（2026-03-08）
- [Don't Fight Backprop: Goodfire's Vision for Intentional Design, w/ Dan Balsam & Tom McGrath](https://cogrev.qiaomu.ai/ep/324-don-t-fight-backprop-goodfire-s-vision-for-intentional-desig-q5IIN0.html)（2026-03-05）
- [Situational Awareness in Government, with UK AISI Chief Scientist Geoffrey Irving](https://cogrev.qiaomu.ai/ep/323-situational-awareness-in-government-with-uk-aisi-chief-scien-M5Ho6A.html)（2026-03-01）
- [Universal Medical Intelligence: OpenAI's Plan to Elevate Human Health, with Karan Singhal](https://cogrev.qiaomu.ai/ep/322-universal-medical-intelligence-openai-s-plan-to-elevate-huma-1YmJoq.html)（2026-02-25）
- [Intelligence with Everyone: RL @ MiniMax, with Olive Song, from AIE NYC & Inference by Turing Post](https://cogrev.qiaomu.ai/ep/321-intelligence-with-everyone-rl-minimax-with-olive-song-from-a-P-1MJC.html)（2026-02-22）
- [Mathematical Superintelligence: Harmonic's Vlad Tenev & Tudor Achim on IMO Gold & Theories of Everything](https://cogrev.qiaomu.ai/ep/320-mathematical-superintelligence-harmonic-s-vlad-tenev-tudor-a--01-Mc.html)（2026-02-18）
- [Approaching the AI Event Horizon? Part 2, w/ Abhi Mahajan, Helen Toner, Jeremie Harris, @8teAPi](https://cogrev.qiaomu.ai/ep/319-approaching-the-ai-event-horizon-part-2-w-abhi-mahajan-helen-c4RR-v.html)（2026-02-14）
- [Approaching the AI Event Horizon? Part 1, w/ James Zou, Sam Hammond, Shoshannah Tekofsky, @8teAPi](https://cogrev.qiaomu.ai/ep/318-approaching-the-ai-event-horizon-part-1-w-james-zou-sam-hamm-1V11Ki.html)（2026-02-13）
- [AGI-Pilled Cyber Defense: Automating Digital Forensics w/ Asymmetric Security Founder Alexis Carlier](https://cogrev.qiaomu.ai/ep/317-agi-pilled-cyber-defense-automating-digital-forensics-w-asym-MwXVLZ.html)（2026-02-08）
- [Infinite Code Context: AI Coding at Enterprise Scale w/ Blitzy CEO Brian Elliott & CTO Sid Pardeshi](https://cogrev.qiaomu.ai/ep/316-infinite-code-context-ai-coding-at-enterprise-scale-w-blitzy-PfN3Hv.html)（2026-02-05）
- [The AI-Powered Biohub: Why Mark Zuckerberg & Priscilla Chan are Investing in Data, from Latent.Space](https://cogrev.qiaomu.ai/ep/315-the-ai-powered-biohub-why-mark-zuckerberg-priscilla-chan-are-1jZpFr.html)（2026-02-01）
- [AI & The Law: Changing Practice, Claude Constitution, & New Rights, w/ Kevin & Alan of Scaling Laws](https://cogrev.qiaomu.ai/ep/314-ai-the-law-changing-practice-claude-constitution-new-rights-26Vhss.html)（2026-01-29）
- [The Internet Computer: Caffeine.ai CEO Dominic Williams on Unstoppable, Self-Writing Software](https://cogrev.qiaomu.ai/ep/313-the-internet-computer-caffeine-ai-ceo-dominic-williams-on-un-LX_iU3.html)（2026-01-25）
- [AMA Part 2: Is Fine-Tuning Dead? How Am I Preparing for AGI? Are We Headed for UBI? & More!](https://cogrev.qiaomu.ai/ep/312-ama-part-2-is-fine-tuning-dead-how-am-i-preparing-for-agi-ar-wlHK2K.html)（2026-01-22）
- [Pioneering PAI: How Daniel Miessler's Personal AI Infrastructure Activates Human Agency & Creativity](https://cogrev.qiaomu.ai/ep/311-pioneering-pai-how-daniel-miessler-s-personal-ai-infrastruct-DbNUDM.html)（2026-01-18）
- [Snowflake VP of AI Baris Gultekin on Bringing AI to Data, Agent Design, Text-2-SQL, RAG & More](https://cogrev.qiaomu.ai/ep/310-snowflake-vp-of-ai-baris-gultekin-on-bringing-ai-to-data-age-txYahw.html)（2026-01-14）
- [AMA Part 1: Is Claude Code AGI?  Are we in a bubble?  Plus Live Player Analysis](https://cogrev.qiaomu.ai/ep/309-ama-part-1-is-claude-code-agi-are-we-in-a-bubble-plus-live-p-NFsnVP.html)（2026-01-09）
- [Building & Scaling the AI Safety Research Community, with Ryan Kidd of MATS](https://cogrev.qiaomu.ai/ep/308-building-scaling-the-ai-safety-research-community-with-ryan-q5A8ne.html)（2026-01-04）
- [Confronting the Intelligence Curse, w/ Luke Drago of Workshop Labs, from the FLI Podcast](https://cogrev.qiaomu.ai/ep/307-confronting-the-intelligence-curse-w-luke-drago-of-workshop-n-tL4H.html)（2026-01-01）
- [Controlling Tools or Aligning Creatures? Emmett Shear (Softmax) & Séb Krier (GDM), from a16z Show](https://cogrev.qiaomu.ai/ep/306-controlling-tools-or-aligning-creatures-emmett-shear-softmax--4G5zF.html)（2025-12-27）
- [The Great Security Update: AI ∧ Formal Methods with Kathleen Fisher of RAND & Byron Cook of AWS](https://cogrev.qiaomu.ai/ep/305-the-great-security-update-ai-formal-methods-with-kathleen-fi-Dm3DHP.html)（2025-12-24）
- [AI 2025 → 2026 Live Show | Part 2](https://cogrev.qiaomu.ai/ep/304-ai-2025-2026-live-show-part-2-D_J2UY.html)（2025-12-19）
- [AI 2025 → 2026 Live Show | Part 1](https://cogrev.qiaomu.ai/ep/303-ai-2025-2026-live-show-part-1-zO38Ok.html)（2025-12-18）
- [AI's Energy & Water Demands: Sorting Fact from Fiction with Andy Masley](https://cogrev.qiaomu.ai/ep/302-ai-s-energy-water-demands-sorting-fact-from-fiction-with-and-zth0v4.html)（2025-12-14）
- [Superintelligence: To Ban or Not to Ban? Max Tegmark & Dean Ball join Liron Shapira on Doom Debates](https://cogrev.qiaomu.ai/ep/301-superintelligence-to-ban-or-not-to-ban-max-tegmark-dean-ball-9O0djo.html)（2025-12-10）
- [Sovereign AI in Poland: Language Adaptation, Local Control & Cost Advantages with Marek Kozlowski](https://cogrev.qiaomu.ai/ep/300-sovereign-ai-in-poland-language-adaptation-local-control-cos-KVos1k.html)（2025-12-06）
- [China's AI Upstarts: How Z.ai Builds, Benchmarks & Ships in Hours, from ChinaTalk](https://cogrev.qiaomu.ai/ep/299-china-s-ai-upstarts-how-z-ai-builds-benchmarks-ships-in-hour-5eMuRt.html)（2025-12-03）
- [Underwriting Superintelligence: How AIUC is using Insurance, Standards, and Audits to Accelerate Adoption while Minimizing Risks](https://cogrev.qiaomu.ai/ep/298-underwriting-superintelligence-how-aiuc-is-using-insurance-s-QigrUV.html)（2025-11-30）
- [AI-Led Sales: How 1Mind's Superhumans Drive Exponential Growth, from the Agents of Scale Podcast](https://cogrev.qiaomu.ai/ep/297-ai-led-sales-how-1mind-s-superhumans-drive-exponential-growt-pMFUmQ.html)（2025-11-27）
- [What AI Means for Students & Teachers: My Keynote from the Michigan Virtual AI Summit](https://cogrev.qiaomu.ai/ep/296-what-ai-means-for-students-teachers-my-keynote-from-the-mich-w1Mj_s.html)（2025-11-25）
- [Escaping AI Slop: How Atlassian Gives AI Teammates Taste, Knowledge, & Workflows, w- Sherif Mansour](https://cogrev.qiaomu.ai/ep/295-escaping-ai-slop-how-atlassian-gives-ai-teammates-taste-know-ut83GE.html)（2025-11-20）
- [AI in the Cancer Journey: How I'm Using AI to Help My Son](https://cogrev.qiaomu.ai/ep/294-ai-in-the-cancer-journey-how-i-m-using-ai-to-help-my-son-cf020d.html)（2025-11-17）
- [My Positive Vision for the AI Future, from the Existential Hope Podcast](https://cogrev.qiaomu.ai/ep/293-my-positive-vision-for-the-ai-future-from-the-existential-ho-DFtg29.html)（2025-11-12）
- [What is Catholic AI? Technology Meets Theology, with Matthew Harvey Sanders, CEO of Longbeard](https://cogrev.qiaomu.ai/ep/292-what-is-catholic-ai-technology-meets-theology-with-matthew-h-MklFfK.html)（2025-11-08）
- [More Truthful AIs Report Conscious Experience: New Mechanistic Research w- Cameron Berg @ AE Studio](https://cogrev.qiaomu.ai/ep/291-more-truthful-ais-report-conscious-experience-new-mechanisti-4Hk_Je.html)（2025-11-05）
- [A Simulationship with AI, with *smiles and kisses you* Star Chris Stokes & Director Bryan Carberry](https://cogrev.qiaomu.ai/ep/290-a-simulationship-with-ai-with-smiles-and-kisses-you-star-chr-ZoMMWz.html)（2025-11-02）
- [Is AI Stalling Out? Cutting Through Capabilities Confusion, w/ Erik Torenberg, from the a16z Podcast](https://cogrev.qiaomu.ai/ep/289-is-ai-stalling-out-cutting-through-capabilities-confusion-w-EPLfU3.html)（2025-10-29）
- [A Positive Vision for the Future: Part 2 with Illia Polosukhin of NEAR](https://cogrev.qiaomu.ai/ep/288-a-positive-vision-for-the-future-part-2-with-illia-polosukhi-1PMyud.html)（2025-10-25）
- [Always Bet on the Models: How Tasklet Puts the Agency in Agents, with CEO Andrew Lee](https://cogrev.qiaomu.ai/ep/287-always-bet-on-the-models-how-tasklet-puts-the-agency-in-agen-ILMVUS.html)（2025-10-22）
- [Chinese AI – They're Just Like Us? With Beijing-Based Concordia AI CEO Brian Tse](https://cogrev.qiaomu.ai/ep/286-chinese-ai-they-re-just-like-us-with-beijing-based-concordia-eNsusQ.html)（2025-10-18）
- [AI Discovered Antibiotics: How Small Data & Small GNNs Led to Big Results, w/ MIT Prof. Jim Collins](https://cogrev.qiaomu.ai/ep/285-ai-discovered-antibiotics-how-small-data-small-gnns-led-to-b-x6WKR7.html)（2025-10-14）
- [Education in the AI Age: a Teacher Rethinks Learning & Purpose, w/ Johan Falk of Graspable AI](https://cogrev.qiaomu.ai/ep/284-education-in-the-ai-age-a-teacher-rethinks-learning-purpose-6uO22D.html)（2025-10-08）
- [The Customer Service Revolution: Building Fin, with Eoghan McCabe & Fergal Reid of Intercom](https://cogrev.qiaomu.ai/ep/283-the-customer-service-revolution-building-fin-with-eoghan-mcc-y2eRWk.html)（2025-10-05）
- [Training an AI Scientist with Feedback from Reality, w- Liam Fedus & Ekin Dogus Cubuk (from a16z)](https://cogrev.qiaomu.ai/ep/282-training-an-ai-scientist-with-feedback-from-reality-w-liam-f-HsDx8p.html)（2025-10-02）
- [The Machines Are Taking Our Jobs - Thank God? Emad Mostaque’s Guide to the next 1000 Days](https://cogrev.qiaomu.ai/ep/281-the-machines-are-taking-our-jobs-thank-god-emad-mostaques-gu--2NpyI.html)（2025-09-28）
- [Stripe's Payments Foundation Model: How Data & Infra Create Compounding Advantage, w/ Emily Sands](https://cogrev.qiaomu.ai/ep/280-stripe-s-payments-foundation-model-how-data-infra-create-com-zhUriz.html)（2025-09-25）
- [Full-Stack AI Safety: Why Defense-in-Depth Might Work, with Far.AI CEO Adam Gleave](https://cogrev.qiaomu.ai/ep/279-full-stack-ai-safety-why-defense-in-depth-might-work-with-fa-iFfTLO.html)（2025-09-20）
- [Can We Stop AI Deception? Apollo Research Tests OpenAI's Deliberative Alignment, w/ Marius Hobbhahn](https://cogrev.qiaomu.ai/ep/278-can-we-stop-ai-deception-apollo-research-tests-openai-s-deli-I3ivZa.html)（2025-09-18）
- [User-Owned AI: On-Chain Training, Inference, and Agents, with NEAR's Illia Polosukhin](https://cogrev.qiaomu.ai/ep/277-user-owned-ai-on-chain-training-inference-and-agents-with-ne-2cdQCT.html)（2025-09-13）
- [Coaching the Creators: Inside the Minds Building Frontier AI with Executive Coach Joe Hudson](https://cogrev.qiaomu.ai/ep/276-coaching-the-creators-inside-the-minds-building-frontier-ai-TX9saS.html)（2025-09-11）
- [Zvi Mowshowitz on Longer Timelines, RL-induced Doom, and Why China is Refusing H20s](https://cogrev.qiaomu.ai/ep/275-zvi-mowshowitz-on-longer-timelines-rl-induced-doom-and-why-c--S6aOh.html)（2025-09-06）
- [In-AI Advertising: Better Answers for Users, Big Questions for Society, with ZeroClick's Ryan Hudson](https://cogrev.qiaomu.ai/ep/274-in-ai-advertising-better-answers-for-users-big-questions-for-fSSwao.html)（2025-09-03）
- [Material Abundance: Radical AI’s Closed-Loop Lab Automates Scientific Discovery](https://cogrev.qiaomu.ai/ep/273-material-abundance-radical-ais-closed-loop-lab-automates-sci-395Aa3.html)（2025-08-30）
- [Untangling Neural Network Mechanisms: Goodfire's Lee Sharkey on Parameter-based Interpretability](https://cogrev.qiaomu.ai/ep/272-untangling-neural-network-mechanisms-goodfire-s-lee-sharkey-tbc5Qf.html)（2025-08-27）
- [What if Humans Weaponize Superintelligence, w/ Tom Davidson, from Future of Life Institute Podcast](https://cogrev.qiaomu.ai/ep/271-what-if-humans-weaponize-superintelligence-w-tom-davidson-fr-_vc6q9.html)（2025-08-23）
- [The AI Whistleblower Initiative: Supporting AGI Insiders When It Matters Most, w/ founder Karl Koch](https://cogrev.qiaomu.ai/ep/270-the-ai-whistleblower-initiative-supporting-agi-insiders-when-lhN8_9.html)（2025-08-20）
- [Autonomous Organizations: Vending Bench & Beyond, w/ Lukas Petersson & Axel Backlund of Andon Labs](https://cogrev.qiaomu.ai/ep/269-autonomous-organizations-vending-bench-beyond-w-lukas-peters-ZAimco.html)（2025-08-16）
- [Dean W. Ball on America's AI Action Plan & 4 Months at the White House](https://cogrev.qiaomu.ai/ep/268-dean-w-ball-on-america-s-ai-action-plan-4-months-at-the-whit-mX4kKQ.html)（2025-08-13）
- [Jake Sullivan on Navigating AI Uncertainty and Managing Competition with China](https://cogrev.qiaomu.ai/ep/267-jake-sullivan-on-navigating-ai-uncertainty-and-managing-comp-3jM4Ni.html)（2025-08-09）
- [Sovereign AI: Geopolitical Strategy & Industrial Policy for Countries 3-193, with Anjney Midha, a16z](https://cogrev.qiaomu.ai/ep/266-sovereign-ai-geopolitical-strategy-industrial-policy-for-cou-UmM7cw.html)（2025-08-06）
- [The Revolution in Services: General Catalyst's AI Rollup Strategy, with Marc Bhargava](https://cogrev.qiaomu.ai/ep/265-the-revolution-in-services-general-catalyst-s-ai-rollup-stra-gT-qbK.html)（2025-08-02）
- [Liability for AI Harms: How Ancient Law Can Govern Frontier Technology Risk, with Prof Gabriel Weil](https://cogrev.qiaomu.ai/ep/264-liability-for-ai-harms-how-ancient-law-can-govern-frontier-t-PwWl2H.html)（2025-07-26）
- [Private Governance: Creating a Market in AI Regulation, with Dr. Gillian Hadfield & Andrew Freedman](https://cogrev.qiaomu.ai/ep/263-private-governance-creating-a-market-in-ai-regulation-with-d-Sqy_Qs.html)（2025-07-22）
- [The Dawn of Dynamic AI: RFT Comes Online, w/ Predibase CEO Dev Rishi, from Inference by Turing Post](https://cogrev.qiaomu.ai/ep/262-the-dawn-of-dynamic-ai-rft-comes-online-w-predibase-ceo-dev-dyCGSY.html)（2025-07-16）
- [Cracking the Medical Code: Why Cleveland Clinic Doctors Love Their Ambience Healthcare AI Scribe](https://cogrev.qiaomu.ai/ep/261-cracking-the-medical-code-why-cleveland-clinic-doctors-love-hyQ_yG.html)（2025-07-10）
- [The Data Factory: Inside the $100B Race for Post-Training Supremacy, with Labelbox CEO Manu Sharma](https://cogrev.qiaomu.ai/ep/260-the-data-factory-inside-the-100b-race-for-post-training-supr-Kr3vNR.html)（2025-07-08）
- [Cheat on Everything: Cluely's Vision for Always-On AI Assistance](https://cogrev.qiaomu.ai/ep/259-cheat-on-everything-cluely-s-vision-for-always-on-ai-assista-jJmndz.html)（2025-07-05）
- [The AI Village: Previewing the Giga-Agent Future with Adam Binksmith, Founder of AI Digest](https://cogrev.qiaomu.ai/ep/258-the-ai-village-previewing-the-giga-agent-future-with-adam-bi-LrYmKD.html)（2025-07-02）
- [402 Payment Required: a New Way for AI Agents to Pay, with Nemil Dalal, Dev Platform Lead @ Coinbase](https://cogrev.qiaomu.ai/ep/257-402-payment-required-a-new-way-for-ai-agents-to-pay-with-nem-P03BXU.html)（2025-06-28）
- [2-Sigma in 2 Hours: How Alpha Schools are Using AI to Revolutionize Education](https://cogrev.qiaomu.ai/ep/256-2-sigma-in-2-hours-how-alpha-schools-are-using-ai-to-revolut-h-KR1G.html)（2025-06-25）
- [Living Lindy: a No-BS Conversation on AI Agents with Flo Crivello](https://cogrev.qiaomu.ai/ep/255-living-lindy-a-no-bs-conversation-on-ai-agents-with-flo-criv-rRwNn-.html)（2025-06-21）
- [Embryology of AI: How Training Data Shapes AI Development w/ Timaeus' Jesse Hoogland & Daniel Murfet](https://cogrev.qiaomu.ai/ep/254-embryology-of-ai-how-training-data-shapes-ai-development-w-t-Cy0m-J.html)（2025-06-19）
- [AI Scouting Report: AI Agents -vs- Agentic AI, from Imagine AI Live](https://cogrev.qiaomu.ai/ep/253-ai-scouting-report-ai-agents-vs-agentic-ai-from-imagine-ai-l-ZMuUcX.html)（2025-06-14）
- [The Decade of May 15-22, 2025: Google's 50X AI Growth & Transformation with Logan Kilpatrick](https://cogrev.qiaomu.ai/ep/252-the-decade-of-may-15-22-2025-google-s-50x-ai-growth-transfor-kp9afm.html)（2025-06-12）
- [a16z on Protecting Little Tech: The Techno-Optimist AI Policy Agenda with Matt Perault, Head of AI Policy](https://cogrev.qiaomu.ai/ep/251-a16z-on-protecting-little-tech-the-techno-optimist-ai-policy-krvJyO.html)（2025-06-09）
- [Don't Make Mirror Life: Synthetic Biologist Kate Adamala on Risks & Responsibility](https://cogrev.qiaomu.ai/ep/250-don-t-make-mirror-life-synthetic-biologist-kate-adamala-on-r-f7MYaV.html)（2025-06-07）
- [What did Google's AI Co-Scientist "Discover"?  The Human Scientists' POV, from the Podovirus podcast](https://cogrev.qiaomu.ai/ep/249-what-did-google-s-ai-co-scientist-discover-the-human-scienti-wLO6SU.html)（2025-06-04）
- [Mechanistic Interpretability: Philosophy, Practice & Progress with Goodfire's Dan Balsam & Tom McGrath](https://cogrev.qiaomu.ai/ep/248-mechanistic-interpretability-philosophy-practice-progress-wi-Ap8YSy.html)（2025-05-29）
- [The Perfect Substrate for AGI, with Replit CEO Amjad Masad](https://cogrev.qiaomu.ai/ep/247-the-perfect-substrate-for-agi-with-replit-ceo-amjad-masad-XlHspW.html)（2025-05-22）
- [The RAISE Act: Minimum Standards for Frontier AI Development, with NY Assembly Member Alex Bores](https://cogrev.qiaomu.ai/ep/246-the-raise-act-minimum-standards-for-frontier-ai-development-7Fdqn4.html)（2025-05-20）
- [Gemini Robotics – AI for the Physical World, with Keerthana Gopalakrishnan and Ted Xiao of Google DeepMind](https://cogrev.qiaomu.ai/ep/245-gemini-robotics-ai-for-the-physical-world-with-keerthana-gop-8burcV.html)（2025-05-17）
- [Titans: Neural Long-Term Memory for LLMs, with author Ali Behrouz](https://cogrev.qiaomu.ai/ep/244-titans-neural-long-term-memory-for-llms-with-author-ali-behr-ShYJc3.html)（2025-05-15）
- [Luma Labs' Diffusion Revolution: from Dream Machine to Multimodal Worldsim - Amit Jain, Jiaming Song](https://cogrev.qiaomu.ai/ep/243-luma-labs-diffusion-revolution-from-dream-machine-to-multimo--wyG1V.html)（2025-05-11）
- [OpenAI's Identity Crisis: History, Culture & Non-Profit Control with ex-employee Steven Adler](https://cogrev.qiaomu.ai/ep/242-openai-s-identity-crisis-history-culture-non-profit-control-ndBcmn.html)（2025-05-08）
- [AI Control: Using Untrusted Systems Safely with Buck Shlegeris of Redwood Research, from the 80,000 Hours Podcast](https://cogrev.qiaomu.ai/ep/241-ai-control-using-untrusted-systems-safely-with-buck-shlegeri-AQgxuI.html)（2025-05-04）
- [Blueprint for AI Armageddon: Josh Clymer Imagines AI Takeover, from the Audio Tokens Podcast](https://cogrev.qiaomu.ai/ep/240-blueprint-for-ai-armageddon-josh-clymer-imagines-ai-takeover-Ndaw_m.html)（2025-05-01）
- [Fiverr Goes All-In on AI: Empowering Creators, Not Replacing Them, with Micha Kaufman, CEO of Fiverr](https://cogrev.qiaomu.ai/ep/239-fiverr-goes-all-in-on-ai-empowering-creators-not-replacing-t-HBf8AR.html)（2025-04-26）
- [Securing Superintelligence: National Security, Espionage & AI Control with Jeremie & Edouard Harris](https://cogrev.qiaomu.ai/ep/238-securing-superintelligence-national-security-espionage-ai-co-tEOeZq.html)（2025-04-23）
- [Is OpenAI's o3 AGI? Zvi Mowshowitz on Early AI Takeoff, the Mechanize launch, Live Players, & Why p(doom) is Rising](https://cogrev.qiaomu.ai/ep/237-is-openai-s-o3-agi-zvi-mowshowitz-on-early-ai-takeoff-the-me-IGyrP_.html)（2025-04-21）
- [AI News Crossover: A Candid Chat with Liron Shapira of Doom Debates](https://cogrev.qiaomu.ai/ep/236-ai-news-crossover-a-candid-chat-with-liron-shapira-of-doom-d-u_O3bh.html)（2025-04-19）
- [Helen Toner: OpenAI Reflections, Adaptation Buffers, and AI in Warfare](https://cogrev.qiaomu.ai/ep/235-helen-toner-openai-reflections-adaptation-buffers-and-ai-in-fmEpVL.html)（2025-04-16）
- [Is a US-China Thucydides Trap Unavoidable? With David C. Kang from the ChinaTalk Podcast](https://cogrev.qiaomu.ai/ep/234-is-a-us-china-thucydides-trap-unavoidable-with-david-c-kang-dbC74J.html)（2025-04-13）
- [New in Nature: Google Agents Beat Human Doctors, Make Scientific Discoveries – With Vivek Natarajan and Anil Palepu](https://cogrev.qiaomu.ai/ep/233-new-in-nature-google-agents-beat-human-doctors-make-scientif-Jo31LE.html)（2025-04-10）
- [Scaling "Thinking": Gemini 2.5 Tech Lead Jack Rae on Reasoning, Long Context, & the Path to AGI](https://cogrev.qiaomu.ai/ep/232-scaling-thinking-gemini-2-5-tech-lead-jack-rae-on-reasoning-u0iIPx.html)（2025-04-05）
- [Reward Hacking by Reasoning Models & Loss of Control Scenarios w/ Jeffrey Ladish of Palisade Research, from FLI Podcast](https://cogrev.qiaomu.ai/ep/231-reward-hacking-by-reasoning-models-loss-of-control-scenarios-Jcfy9D.html)（2025-04-02）
- [Shortwave Rides the Tidal Wave: Inbox Agents, Hyper-Growth & Hiring AI Managers, with CEO Andrew Lee](https://cogrev.qiaomu.ai/ep/230-shortwave-rides-the-tidal-wave-inbox-agents-hyper-growth-hir-b51Nwe.html)（2025-03-29）
- [Code Context is King: Augment’s AI Assistant for Professional Software Engineers, with Guy Gur-Ari](https://cogrev.qiaomu.ai/ep/229-code-context-is-king-augments-ai-assistant-for-professional-LYWkB6.html)（2025-03-27）
- [Unlocking Cells' Secrets: Diffusion, Deconvolution, & Discovery with Siyu He, author of Squidiff & CORAL](https://cogrev.qiaomu.ai/ep/228-unlocking-cells-secrets-diffusion-deconvolution-discovery-wi-tmyuSy.html)（2025-03-23）
- [a16z on AI Voices: Call Centers, Coaches, and Companions with Olivia Moore & Anish Acharya](https://cogrev.qiaomu.ai/ep/227-a16z-on-ai-voices-call-centers-coaches-and-companions-with-o-q3G6bT.html)（2025-03-19）
- [Agency over AI? Allan Dafoe on Technological Determinism & DeepMind's Safety Plans, from 80000 Hours](https://cogrev.qiaomu.ai/ep/226-agency-over-ai-allan-dafoe-on-technological-determinism-deep-lLwqp3.html)（2025-03-15）
- [China's Tech Tightrope: Power, Regulation, and the AI Race with Angela Zhang](https://cogrev.qiaomu.ai/ep/225-china-s-tech-tightrope-power-regulation-and-the-ai-race-with-GD3pZc.html)（2025-03-12）
- [Historic AI Developments & the Emerging Shape of Superintelligence, from the Consistently Candid Podcast](https://cogrev.qiaomu.ai/ep/224-historic-ai-developments-the-emerging-shape-of-superintellig-5Ql6Jj.html)（2025-03-08）
- [Frontier Models for Frontier Science with Professor Derya Unutmaz, Immunologist & ChatGPT Pro Grantee](https://cogrev.qiaomu.ai/ep/223-frontier-models-for-frontier-science-with-professor-derya-un-JOKvE4.html)（2025-03-06）
- [US-China Relations: History, Culture, and AI Competition, with Noah Smith, from Econ 102](https://cogrev.qiaomu.ai/ep/222-us-china-relations-history-culture-and-ai-competition-with-n-xxuqNt.html)（2025-03-01）
- [The Adversarial Mind: Defeating AI Defenses with Nicholas Carlini of Google DeepMind](https://cogrev.qiaomu.ai/ep/221-the-adversarial-mind-defeating-ai-defenses-with-nicholas-car-fyGdln.html)（2025-02-27）
- [New Jersey’s AI Moonshot: Governor Phil Murphy on Partnerships, Progress, and Preparedness](https://cogrev.qiaomu.ai/ep/220-new-jerseys-ai-moonshot-governor-phil-murphy-on-partnerships-VRULY6.html)（2025-02-22）
- [Inference Scaling, Alignment Faking, Deal Making? Frontier Research with Ryan Greenblatt of Redwood Research](https://cogrev.qiaomu.ai/ep/219-inference-scaling-alignment-faking-deal-making-frontier-rese-KlXTH7.html)（2025-02-20）
- [An Application-Free Future? Speaking Directly to Data with illumex CEO Inna Tokarev Sela](https://cogrev.qiaomu.ai/ep/218-an-application-free-future-speaking-directly-to-data-with-il--h9omY.html)（2025-02-15）
- [Claude Cooperates! Exploring Cultural Evolution in LLM Societies, with Aron Vallinder & Edward Hughes](https://cogrev.qiaomu.ai/ep/217-claude-cooperates-exploring-cultural-evolution-in-llm-societ-T_guv5.html)（2025-02-12）
- [Software Supernova: Lovable's "Superhuman Full Stack Engineer" to Transform Idea to App in Seconds](https://cogrev.qiaomu.ai/ep/216-software-supernova-lovable-s-superhuman-full-stack-engineer-SM9_FS.html)（2025-02-09）
- [Software Supernova: Bolt.new - The AI Web App Developer In Your Browser](https://cogrev.qiaomu.ai/ep/215-software-supernova-bolt-new-the-ai-web-app-developer-in-your-8WpnDI.html)（2025-02-09）
- [Gemini's Next Frontier: 2.0 Flash, Flash Lite Strategy & Real-Time APIs with Logan K from Google Deepmind](https://cogrev.qiaomu.ai/ep/214-gemini-s-next-frontier-2-0-flash-flash-lite-strategy-real-ti-QJ1y19.html)（2025-02-06）
- [Distributed Training, Decentralized AI: Prime Intellect's Master Plan to Make AI Too Cheap to Meter](https://cogrev.qiaomu.ai/ep/213-distributed-training-decentralized-ai-prime-intellect-s-mast-atJo38.html)（2025-02-05）
- [Aaron Levie, CEO of Box, on Box AI, Enterprise Enthusiasm, and the Evolution of SaaS](https://cogrev.qiaomu.ai/ep/212-aaron-levie-ceo-of-box-on-box-ai-enterprise-enthusiasm-and-t-YAHXJ-.html)（2025-01-29）
- [Emergency Pod: Reinforcement Learning Works! Reflecting on Chinese Reasoning Models DeepSeek-R1 and Kimi k1.5](https://cogrev.qiaomu.ai/ep/211-emergency-pod-reinforcement-learning-works-reflecting-on-chi-MbX9J1.html)（2025-01-25）
- [Material Progress: Developing AI's Scientific Intuition, with Orbital Materials' Jonathan Godwin & Tim Duignan](https://cogrev.qiaomu.ai/ep/210-material-progress-developing-ai-s-scientific-intuition-with-eVlXEJ.html)（2025-01-22）
- [Dodging Latent Space Detectors: Obfuscated Activation Attacks with Luke, Erik, and Scott.](https://cogrev.qiaomu.ai/ep/209-dodging-latent-space-detectors-obfuscated-activation-attacks-WirqcB.html)（2025-01-18）
- [Gene Hunting with o1-pro: Reasoning about Rare Diseases with ChatGPT Pro Grantee Dr. Catherine Brownstein](https://cogrev.qiaomu.ai/ep/208-gene-hunting-with-o1-pro-reasoning-about-rare-diseases-with-IM1-Vt.html)（2025-01-15）
- [AI AMA – Part 2: AI Utopia, Consciousness, and the Future of Work](https://cogrev.qiaomu.ai/ep/207-ai-ama-part-2-ai-utopia-consciousness-and-the-future-of-work-pQFhhy.html)（2025-01-08）
- [AI AMA – Part 1: OpenAI’s o3, Deliberative Alignment, and AI Surprises of 2024](https://cogrev.qiaomu.ai/ep/206-ai-ama-part-1-openais-o3-deliberative-alignment-and-ai-surpr-Eifck_.html)（2025-01-07）
- [Teaching AI to See: A Technical Deep-Dive on Vision Language Models with Will Hardman of Veratai](https://cogrev.qiaomu.ai/ep/205-teaching-ai-to-see-a-technical-deep-dive-on-vision-language-mLgyZ5.html)（2025-01-03）
- [roon's Heroic Duty: Will "the Good Guys" Build AGI First?  (from Doom Debates)](https://cogrev.qiaomu.ai/ep/204-roon-s-heroic-duty-will-the-good-guys-build-agi-first-from-d-RIOQXI.html)（2024-12-28）
- [Emad Mostaque on the Intelligent Internet and Universal Basic AI](https://cogrev.qiaomu.ai/ep/203-emad-mostaque-on-the-intelligent-internet-and-universal-basi-SEd3hz.html)（2024-12-25）
- [Can AIs do AI R&D? Reviewing REBench Results with Neev Parikh of METR](https://cogrev.qiaomu.ai/ep/202-can-ais-do-ai-r-d-reviewing-rebench-results-with-neev-parikh-SX8Mxy.html)（2024-12-21）
- [Breakthroughs in AI for Biology: AI Lab Groups & Protein Model Interpretability with Prof James Zou](https://cogrev.qiaomu.ai/ep/201-breakthroughs-in-ai-for-biology-ai-lab-groups-protein-model-plq7R9.html)（2024-12-18）
- [Scouting Frontiers in AI for Biology: Dynamics, Diffusion, and Design, with Amelie Schreiber](https://cogrev.qiaomu.ai/ep/200-scouting-frontiers-in-ai-for-biology-dynamics-diffusion-and-enNrsT.html)（2024-12-14）
- [Building Government's Largest Civilian AI Team with DHS AI Corps' Director, Michael Boyce](https://cogrev.qiaomu.ai/ep/199-building-government-s-largest-civilian-ai-team-with-dhs-ai-c-9ElIYR.html)（2024-12-11）
- [Emergency Pod: o1 Schemes Against Users, with Alexander Meinke from Apollo Research](https://cogrev.qiaomu.ai/ep/198-emergency-pod-o1-schemes-against-users-with-alexander-meinke-pB3gvX.html)（2024-12-07）
- [Automating Scientific Discovery, with Andrew White, Head of Science at Future House](https://cogrev.qiaomu.ai/ep/197-automating-scientific-discovery-with-andrew-white-head-of-sc-umFbsC.html)（2024-12-05）
- [The Evolution of AI Agents: Lessons from 2024, with MultiOn CEO Div Garg](https://cogrev.qiaomu.ai/ep/196-the-evolution-of-ai-agents-lessons-from-2024-with-multion-ce-30Nf4x.html)（2024-12-03）
- [Beyond Preference Alignment: Teaching AIs to Play Roles & Respect Norms, with Tan Zhi Xuan](https://cogrev.qiaomu.ai/ep/195-beyond-preference-alignment-teaching-ais-to-play-roles-respe-Y21syy.html)（2024-11-30）
- [Is an AI Arms Race Inevitable? with Robert Wright of Nonzero Newsletter & Podcast](https://cogrev.qiaomu.ai/ep/194-is-an-ai-arms-race-inevitable-with-robert-wright-of-nonzero-u80eXp.html)（2024-11-27）
- [Designing the Future: Inside Canva's AI Strategy with John Milinovich, GenAI Product Lead at Canva](https://cogrev.qiaomu.ai/ep/193-designing-the-future-inside-canva-s-ai-strategy-with-john-mi-onAFbx.html)（2024-11-23）
- [Everything You Wanted to Know About LLM Post-Training, with Nathan Lambert of Allen Institute for AI](https://cogrev.qiaomu.ai/ep/192-everything-you-wanted-to-know-about-llm-post-training-with-n-LVXtFn.html)（2024-11-21）
- [Zvi’s POV: Ilya’s SSI, OpenAI’s o1, Claude Computer Use, Trump’s election, and more](https://cogrev.qiaomu.ai/ep/191-zvis-pov-ilyas-ssi-openais-o1-claude-computer-use-trumps-ele-aiP7z-.html)（2024-11-16）
- [AGI Lab Transparency Requirements & Whistleblower Protections, with Dean W. Ball & Daniel Kokotajlo](https://cogrev.qiaomu.ai/ep/190-agi-lab-transparency-requirements-whistleblower-protections-dolDtu.html)（2024-11-12）
- [AI Under Trump? The Stakes of 2024 w/ Joshua Steinman [Pt 2 of 2]](https://cogrev.qiaomu.ai/ep/189-ai-under-trump-the-stakes-of-2024-w-joshua-steinman-pt-2-of-zU4X6g.html)（2024-11-02）
- [The Case for Trump and the Future of AI – Part 1, with Samuel Hammond, Senior Economist, Foundation of American Innovation](https://cogrev.qiaomu.ai/ep/188-the-case-for-trump-and-the-future-of-ai-part-1-with-samuel-h-D5j87v.html)（2024-11-01）
- [Breaking: Gemini's Major Update - Search, JSON & Code Features Revealed by Google PMs](https://cogrev.qiaomu.ai/ep/187-breaking-gemini-s-major-update-search-json-code-features-rev-R7M785.html)（2024-10-31）
- [Training Zamba: A Hybrid Model Master Class with Zyphra's Quentin Anthony](https://cogrev.qiaomu.ai/ep/186-training-zamba-a-hybrid-model-master-class-with-zyphra-s-que-xFU7mf.html)（2024-10-30）
- [Mind Hacked by AI: A Cautionary Tale, From a LessWrong User's Confession](https://cogrev.qiaomu.ai/ep/185-mind-hacked-by-ai-a-cautionary-tale-from-a-lesswrong-user-s-dja6Vp.html)（2024-10-26）
- [Can AIs Generate Novel Research Ideas? with lead author Chenglei Si](https://cogrev.qiaomu.ai/ep/184-can-ais-generate-novel-research-ideas-with-lead-author-cheng-IOCi6q.html)（2024-10-23）
- [GELU, MMLU, & X-Risk Defense in Depth, with the Great Dan Hendrycks](https://cogrev.qiaomu.ai/ep/183-gelu-mmlu-x-risk-defense-in-depth-with-the-great-dan-hendryc-1fyqkH.html)（2024-10-19）
- [Leading Indicators of AI Danger: Owain Evans on Situational Awareness & Out-of-Context Reasoning, from The Inside View](https://cogrev.qiaomu.ai/ep/182-leading-indicators-of-ai-danger-owain-evans-on-situational-a-Sl0AQI.html)（2024-10-16）
- [Convergent Evolution: The Co-Revolution of AI & Biology with Professor Michael Levin & Staff Scientist Leo Pio Lopez](https://cogrev.qiaomu.ai/ep/181-convergent-evolution-the-co-revolution-of-ai-biology-with-pr-WRtWo-.html)（2024-10-12）
- [Runway's Video Revolution: Empowering Creators with General World Models, with CTO Anastasis Germanidis](https://cogrev.qiaomu.ai/ep/180-runway-s-video-revolution-empowering-creators-with-general-w-tCZd5f.html)（2024-10-09）
- [Biologically Inspired AI Alignment & Neglected Approaches to AI Safety, with Judd Rosenblatt and Mike Vaiana of AE Studio](https://cogrev.qiaomu.ai/ep/179-biologically-inspired-ai-alignment-neglected-approaches-to-a-1C3yLK.html)（2024-10-05）
- [Automating Software Engineering: Genie Tops SWE-Bench, w/ Alistair Pullen, from Latent.Space podcast](https://cogrev.qiaomu.ai/ep/178-automating-software-engineering-genie-tops-swe-bench-w-alist-mij1OI.html)（2024-10-02）
- [Zapier's AI Revolution: From No-Code Pioneer to LLM Knowledge Worker](https://cogrev.qiaomu.ai/ep/177-zapier-s-ai-revolution-from-no-code-pioneer-to-llm-knowledge-6JbKvV.html)（2024-09-26）
- [Anthropic's Responsible Scaling Policy, with Nick Joseph, from the 80,000 Hours Podcast](https://cogrev.qiaomu.ai/ep/176-anthropic-s-responsible-scaling-policy-with-nick-joseph-from-X0wElH.html)（2024-09-25）
- [The Evolution Revolution: Scouting Frontiers in AI for Biology with Brian Hie](https://cogrev.qiaomu.ai/ep/175-the-evolution-revolution-scouting-frontiers-in-ai-for-biolog-1VVAF4.html)（2024-09-20）
- [The Professional Network for AI Agents, with Agent.ai Engineering Lead Andrei Oprisan](https://cogrev.qiaomu.ai/ep/174-the-professional-network-for-ai-agents-with-agent-ai-enginee-48a_aP.html)（2024-09-18）
- [Red Teaming o1 Part 2/2– Detecting Deception with Marius Hobbhahn of Apollo Research](https://cogrev.qiaomu.ai/ep/173-red-teaming-o1-part-2-2-detecting-deception-with-marius-hobb-ObW4Dt.html)（2024-09-14）
- [Red Teaming o1 Part 1/2– Automated Jailbreaking with Haize Labs' Leonard Tang, Aidan Ewart, and Brian Huang](https://cogrev.qiaomu.ai/ep/172-red-teaming-o1-part-1-2-automated-jailbreaking-with-haize-la-98V8wV.html)（2024-09-14）
- [The Path to Utopia, with Nick Bostrom – from Clearer Thinking with Spencer Greenberg](https://cogrev.qiaomu.ai/ep/171-the-path-to-utopia-with-nick-bostrom-from-clearer-thinking-w-HvRc23.html)（2024-09-11）
- [Empathy for AIs: Reframing Alignment with Robopsychologist Yeshua God](https://cogrev.qiaomu.ai/ep/170-empathy-for-ais-reframing-alignment-with-robopsychologist-ye-L4CyBt.html)（2024-09-07）
- [AI Live Players: the Geopolitics & Strategic Dynamics of AI, with Samo Burja of Bismarck Analysis](https://cogrev.qiaomu.ai/ep/169-ai-live-players-the-geopolitics-strategic-dynamics-of-ai-wit-_3z-vq.html)（2024-09-04）
- [Computational Life: How Self-Replicators Arise from Randomness, with Google’s Ettore Randazzo and Luca Versari](https://cogrev.qiaomu.ai/ep/168-computational-life-how-self-replicators-arise-from-randomnes-1kBGDV.html)（2024-08-30）
- [Scaling Forecasting: AI Forecasting Tournaments & the Road to Epistemic Security, with Deger Turan, CEO of Metaculus](https://cogrev.qiaomu.ai/ep/167-scaling-forecasting-ai-forecasting-tournaments-the-road-to-e-mpz1cu.html)（2024-08-28）
- [Final Analysis on CA’s AI Bill SB 1047 with Nathan Calvin, Dean W. Ball, and Steve Newman](https://cogrev.qiaomu.ai/ep/166-final-analysis-on-cas-ai-bill-sb-1047-with-nathan-calvin-dea-G8BXft.html)（2024-08-24）
- [AI Automation: Making AI Work for You](https://cogrev.qiaomu.ai/ep/165-ai-automation-making-ai-work-for-you-mPGkOb.html)（2024-08-21）
- [Popular Mechanistic Interpretability: Goodfire Lights the Way to AI Safety](https://cogrev.qiaomu.ai/ep/164-popular-mechanistic-interpretability-goodfire-lights-the-way-45g68Z.html)（2024-08-17）
- [Understanding AI "Understanding" with Robert Wright of Nonzero Newsletter & Podcast](https://cogrev.qiaomu.ai/ep/163-understanding-ai-understanding-with-robert-wright-of-nonzero-9EdR6A.html)（2024-08-14）
- [AI Consciousness? Exploring the Possibility with Prof. Eric Schwitzgebel](https://cogrev.qiaomu.ai/ep/162-ai-consciousness-exploring-the-possibility-with-prof-eric-sc-iXKoLL.html)（2024-08-10）
- [Self-Driving Cars: Timothy B. Lee Answers All the Questions You Were Too Afraid to Ask](https://cogrev.qiaomu.ai/ep/161-self-driving-cars-timothy-b-lee-answers-all-the-questions-yo-doIcym.html)（2024-08-07）
- [Underlord: Descript’s AI Video Editor, with CEO Andrew Mason](https://cogrev.qiaomu.ai/ep/160-underlord-descripts-ai-video-editor-with-ceo-andrew-mason-oTf5qA.html)（2024-08-03）
- [AI Avatars & the Future of Video, with HeyGen CEO Joshua Xu and Benchmark's Victor Lazarte](https://cogrev.qiaomu.ai/ep/159-ai-avatars-the-future-of-video-with-heygen-ceo-joshua-xu-and-5ShB2P.html)（2024-08-02）
- [Zooming Out on AI, from the Nick Halaris Show](https://cogrev.qiaomu.ai/ep/158-zooming-out-on-ai-from-the-nick-halaris-show-eMD88u.html)（2024-07-27）
- [From Poetry to Programming: The Evolution of Prompt Engineering with Riley Goodside of Scale AI](https://cogrev.qiaomu.ai/ep/157-from-poetry-to-programming-the-evolution-of-prompt-engineeri-wjaSyL.html)（2024-07-24）
- [Long-Term Memory for LLMs, with HippoRAG author Bernal Jiménez Gutierrez](https://cogrev.qiaomu.ai/ep/156-long-term-memory-for-llms-with-hipporag-author-bernal-jimene-YQauz1.html)（2024-07-19）
- [Guaranteed Safe AI? World Models, Safety Specs, & Verifiers, with Nora Ammann & Ben Goldhaber](https://cogrev.qiaomu.ai/ep/155-guaranteed-safe-ai-world-models-safety-specs-verifiers-with-fyVrnb.html)（2024-07-17）
- [The ARC Prize: Efficiency, Intuition, and AGI, with Mike Knoop, co-founder of Zapier](https://cogrev.qiaomu.ai/ep/154-the-arc-prize-efficiency-intuition-and-agi-with-mike-knoop-c-XECXoS.html)（2024-07-13）
- [A Foundation Model for Food, with Chef Robotics CEO Rajat Bhageria](https://cogrev.qiaomu.ai/ep/153-a-foundation-model-for-food-with-chef-robotics-ceo-rajat-bha-ZUTVwg.html)（2024-07-11）
- [Delving into The Prompt Report, with Sander Schulhoff of LearnPrompting.org](https://cogrev.qiaomu.ai/ep/152-delving-into-the-prompt-report-with-sander-schulhoff-of-lear-tv8kqW.html)（2024-07-09）
- [Building an Intelligent Business OS, with Runway CEO Siqi Chen](https://cogrev.qiaomu.ai/ep/151-building-an-intelligent-business-os-with-runway-ceo-siqi-che-KIvlDH.html)（2024-07-06）
- [The State Space Model Revolution, with Albert Gu](https://cogrev.qiaomu.ai/ep/150-the-state-space-model-revolution-with-albert-gu-1zjMal.html)（2024-07-04）
- [AI Safety Regulations: Prudent or Paranoid? with a16z's Martin Casado](https://cogrev.qiaomu.ai/ep/149-ai-safety-regulations-prudent-or-paranoid-with-a16z-s-martin-oZ7788.html)（2024-06-28）
- [Fluid Intelligence: Simulating Solutions with Tim Duignan](https://cogrev.qiaomu.ai/ep/148-fluid-intelligence-simulating-solutions-with-tim-duignan-iESd3e.html)（2024-06-26）
- [The AI Doctor Can See You Now, with Vivek Natarajan and Khaled Saab from Google](https://cogrev.qiaomu.ai/ep/147-the-ai-doctor-can-see-you-now-with-vivek-natarajan-and-khale-n_YEwS.html)（2024-06-22）
- [The Case for Cautious AI Optimism, from the Consistently Candid podcast](https://cogrev.qiaomu.ai/ep/146-the-case-for-cautious-ai-optimism-from-the-consistently-cand-JU0kCR.html)（2024-06-19）
- [Building Brave: Private Search, One AI Layer at a Time with Josep M. Pujol](https://cogrev.qiaomu.ai/ep/145-building-brave-private-search-one-ai-layer-at-a-time-with-jo-S3y1R5.html)（2024-06-15）
- [Investing in AI for Hard Tech, with Eric Vishria of Benchmark and Sergiy Nesterenko of Quilter](https://cogrev.qiaomu.ai/ep/144-investing-in-ai-for-hard-tech-with-eric-vishria-of-benchmark-64iPS9.html)（2024-06-13）
- [Feeling the AGI with Flo Crivello](https://cogrev.qiaomu.ai/ep/143-feeling-the-agi-with-flo-crivello-63Icns.html)（2024-06-08）
- [Inside Google’s AI Studio with Logan Kilpatrick](https://cogrev.qiaomu.ai/ep/142-inside-googles-ai-studio-with-logan-kilpatrick-CD1ku-.html)（2024-06-06）
- [The Implications of AI on the Global Balance of Power with Alex Wang, Andrew Ng, Jack Clark and Cory Booker](https://cogrev.qiaomu.ai/ep/141-the-implications-of-ai-on-the-global-balance-of-power-with-a-SuSQpn.html)（2024-06-04）
- [AI Scouting Report: NOT Investment Advice Edition](https://cogrev.qiaomu.ai/ep/140-ai-scouting-report-not-investment-advice-edition-FUbqQm.html)（2024-06-01）
- [The AI Revolution in Biology: From Vaccines to Protein Engineering with Amelie Schreiber](https://cogrev.qiaomu.ai/ep/139-the-ai-revolution-in-biology-from-vaccines-to-protein-engine-_YEk0Y.html)（2024-05-25）
- [Claude Interpreter: Taking Safe AI to Market with Alex Albert of Anthropic](https://cogrev.qiaomu.ai/ep/138-claude-interpreter-taking-safe-ai-to-market-with-alex-albert-5cQouQ.html)（2024-05-22）
- [OpenAI's Safety Team Exodus: Ilya Departs, Leike Speaks Out, Altman Responds - Zvi Analyzes Fallout](https://cogrev.qiaomu.ai/ep/137-openai-s-safety-team-exodus-ilya-departs-leike-speaks-out-al-lvjs-1.html)（2024-05-19）
- [OpenAI and Google Race to "Her" - Is the Big Tech Singularity Near? Part 1 with Zvi Mowshowitz](https://cogrev.qiaomu.ai/ep/136-openai-and-google-race-to-her-is-the-big-tech-singularity-ne-8fYtGC.html)（2024-05-18）
- [The Open Source AI Question - Part 2 | Robert Wright & Nathan Labenz](https://cogrev.qiaomu.ai/ep/135-the-open-source-ai-question-part-2-robert-wright-nathan-labe-Kn6Rxe.html)（2024-05-15）
- [Governing Frontier AI, with CA Senator Scott Wiener, Author of SB 1047](https://cogrev.qiaomu.ai/ep/134-governing-frontier-ai-with-ca-senator-scott-wiener-author-of-U7INYD.html)（2024-05-09）
- [TinyGPU, Massive Learning, with Adam Majmudar](https://cogrev.qiaomu.ai/ep/133-tinygpu-massive-learning-with-adam-majmudar-WulkHH.html)（2024-05-06）
- [Emergency Pod: Examining SB 1047, California's Proposed AI Legislation](https://cogrev.qiaomu.ai/ep/132-emergency-pod-examining-sb-1047-california-s-proposed-ai-leg-DcQdy8.html)（2024-05-04）
- [Infinite AI Interns: How Young Professionals Can Win in an AI-Powered World](https://cogrev.qiaomu.ai/ep/131-infinite-ai-interns-how-young-professionals-can-win-in-an-ai-KIhGtk.html)（2024-05-03）
- [The State of AI, from the 80,000 Hours Podcast](https://cogrev.qiaomu.ai/ep/130-the-state-of-ai-from-the-80-000-hours-podcast-d7QrAE.html)（2024-05-01）
- [Organic Holodeck with Prophetic AI co-founders Eric Wollberg and Wesley Berry](https://cogrev.qiaomu.ai/ep/129-organic-holodeck-with-prophetic-ai-co-founders-eric-wollberg-WFqTvy.html)（2024-04-29）
- [Merging Man and Machine? The Neurotech Frontier with Dean W. Ball](https://cogrev.qiaomu.ai/ep/128-merging-man-and-machine-the-neurotech-frontier-with-dean-w-b-dRxola.html)（2024-04-26）
- [Automating Circuit Board Design Using Reinforcement Learning w Sergiy Nesterenko, Founder of Quilter on Autopilot](https://cogrev.qiaomu.ai/ep/127-automating-circuit-board-design-using-reinforcement-learning-0M_0xh.html)（2024-04-25）
- [Robotics Research Update, with Keerthana Gopalakrishnan and Ted Xiao of Google DeepMind](https://cogrev.qiaomu.ai/ep/126-robotics-research-update-with-keerthana-gopalakrishnan-and-t-oH2vjG.html)（2024-04-22）
- [AI Inference: Good, Fast, and Cheap, with Lin Qiao & Dmytro Ivchenko of Fireworks AI](https://cogrev.qiaomu.ai/ep/125-ai-inference-good-fast-and-cheap-with-lin-qiao-dmytro-ivchen-KVlEYL.html)（2024-04-20）
- [Reading Minds from Shared Latent Space](https://cogrev.qiaomu.ai/ep/124-reading-minds-from-shared-latent-space-7_BS8t.html)（2024-04-17）
- [Interfacing with AI, with Linus Lee of Notion](https://cogrev.qiaomu.ai/ep/123-interfacing-with-ai-with-linus-lee-of-notion-t8QJEG.html)（2024-04-15）
- [Data, data, everywhere - enough for AGI?](https://cogrev.qiaomu.ai/ep/122-data-data-everywhere-enough-for-agi-IdCWaW.html)（2024-04-13）
- [Inside China's AI Ecosystem: A View From Beijing](https://cogrev.qiaomu.ai/ep/121-inside-china-s-ai-ecosystem-a-view-from-beijing-PQSrMv.html)（2024-04-10）
- [The AI Email Assistant I've Been Waiting for, with Andrew Lee of Shortwave](https://cogrev.qiaomu.ai/ep/120-the-ai-email-assistant-i-ve-been-waiting-for-with-andrew-lee-BeoMJB.html)（2024-04-08）
- [Opening AI's Black Box with Prof. David Bau, Koyena Pal, and Eric Todd of Northeastern University](https://cogrev.qiaomu.ai/ep/119-opening-ai-s-black-box-with-prof-david-bau-koyena-pal-and-er-O4Opvj.html)（2024-04-05）
- [Unbounded AI-Assisted Research with Elicit Founders Andreas Stuhlmüller and Jungwon Byun](https://cogrev.qiaomu.ai/ep/118-unbounded-ai-assisted-research-with-elicit-founders-andreas-sv1E5L.html)（2024-04-03）
- [Mamba-Palooza: 90 Days of Mamba-Inspired Research with Jason Meaux: Part 2](https://cogrev.qiaomu.ai/ep/117-mamba-palooza-90-days-of-mamba-inspired-research-with-jason-MwIiQs.html)（2024-03-30）
- [Mamba-Palooza: 90 Days of Mamba-Inspired Research with Jason Meaux: Part 1](https://cogrev.qiaomu.ai/ep/116-mamba-palooza-90-days-of-mamba-inspired-research-with-jason-Bg1LQ_.html)（2024-03-30）
- [Exploitable by Default: Vulnerabilities in GPT-4 APIs and “Superhuman” Go AIs with Adam Gleave of Far.ai](https://cogrev.qiaomu.ai/ep/115-exploitable-by-default-vulnerabilities-in-gpt-4-apis-and-sup-LKyGiV.html)（2024-03-27）
- [Surveying 2,700+ AI Researchers on the Industry's Future with Katja Grace of AI Impacts](https://cogrev.qiaomu.ai/ep/114-surveying-2-700-ai-researchers-on-the-industry-s-future-with-hliLDN.html)（2024-03-21）
- [Envisioning a Positive AI Future, with Tim Reutemann, Author of Liquid Reign](https://cogrev.qiaomu.ai/ep/113-envisioning-a-positive-ai-future-with-tim-reutemann-author-o-DCWt9V.html)（2024-03-16）
- [AI and the Practice of Law: from CaseText to CoCounsel, with Pablo Arredondo, VP of CoCounsel for Thomson Reuters](https://cogrev.qiaomu.ai/ep/112-ai-and-the-practice-of-law-from-casetext-to-cocounsel-with-p-pfMCtu.html)（2024-03-14）
- [Flow Engineering and Code Integrity at Scale with Itamar Friedman, CEO of Codium AI](https://cogrev.qiaomu.ai/ep/111-flow-engineering-and-code-integrity-at-scale-with-itamar-fri-KPBVND.html)（2024-03-12）
- [The Example Engine: How Exa Is Creating the AI Librarian for the Web with Will Bryk, CEO of Exa](https://cogrev.qiaomu.ai/ep/110-the-example-engine-how-exa-is-creating-the-ai-librarian-for-dT_Ric.html)（2024-03-05）
- [Amazing Answers: Richard Socher on how You.com is Reimagining Search with AI](https://cogrev.qiaomu.ai/ep/109-amazing-answers-richard-socher-on-how-you-com-is-reimagining-twFSPg.html)（2024-02-29）
- [Is AGI Far? With Robin Hanson, Economist at George Mason University and Researcher at Future of Humanity Institute](https://cogrev.qiaomu.ai/ep/108-is-agi-far-with-robin-hanson-economist-at-george-mason-unive-YzTEdO.html)（2024-02-27）
- [Looking Up the AI Exponential with Azeem Azhar of the Exponential View](https://cogrev.qiaomu.ai/ep/107-looking-up-the-ai-exponential-with-azeem-azhar-of-the-expone-OxnR5a.html)（2024-02-23）
- [OpenAI Sora, Google Gemini, and Meta with Zvi Mowshowitz](https://cogrev.qiaomu.ai/ep/106-openai-sora-google-gemini-and-meta-with-zvi-mowshowitz-ra3kpT.html)（2024-02-21）
- [AI Friends, Real Relationships with Eugenia Kuyda, Replika's Founder & CEO](https://cogrev.qiaomu.ai/ep/105-ai-friends-real-relationships-with-eugenia-kuyda-replika-s-f-584wLI.html)（2024-02-16）
- [AI Abundance: a16z Partners Justine Moore and Anish Acharya Forecast a New Era for Consumers](https://cogrev.qiaomu.ai/ep/104-ai-abundance-a16z-partners-justine-moore-and-anish-acharya-f-BfcW4g.html)（2024-02-13）
- [Dr. Michael Levin on Embodied Minds and Cognitive Agents](https://cogrev.qiaomu.ai/ep/103-dr-michael-levin-on-embodied-minds-and-cognitive-agents-LYyGG9.html)（2024-02-09）
- [A Brief History of Biological and Artificial Intelligence with Max Bennett](https://cogrev.qiaomu.ai/ep/102-a-brief-history-of-biological-and-artificial-intelligence-wi-HTvaAv.html)（2024-02-07）
- [AI & Identity, from East & West, with Yohei Nakajima GP at Untapped Capital and BabyAGI Creator](https://cogrev.qiaomu.ai/ep/101-ai-identity-from-east-west-with-yohei-nakajima-gp-at-untappe-LjCTzj.html)（2024-02-01）
- [Ignore Previous Instructions and Listen To This Interview with Sander Schulhoff, CEO of Learnprompting.org](https://cogrev.qiaomu.ai/ep/100-ignore-previous-instructions-and-listen-to-this-interview-wi-W-WyN4.html)（2024-01-30）
- [The Pixel Revolution Part 2 with Suhail Doshi, Founder of Playground AI](https://cogrev.qiaomu.ai/ep/099-the-pixel-revolution-part-2-with-suhail-doshi-founder-of-pla-89obYA.html)（2024-01-26）
- [NYTimes vs OpenAI: Generative AI and the Law with Cecilia Ziniti, Founder and CEO of GC AI](https://cogrev.qiaomu.ai/ep/098-nytimes-vs-openai-generative-ai-and-the-law-with-cecilia-zin-ql_FjO.html)（2024-01-23）
- [The Quest for Autonomous Web Agents with Div Garg, Cofounder and CEO of MultiOn](https://cogrev.qiaomu.ai/ep/097-the-quest-for-autonomous-web-agents-with-div-garg-cofounder--uWnHG.html)（2024-01-20）
- [Unlocking Enterprise Data with Knowledge Graphs and AI, with Juan Sequeda, Principal Scientist and Head of AI Lab at data.world](https://cogrev.qiaomu.ai/ep/096-unlocking-enterprise-data-with-knowledge-graphs-and-ai-with-1Km4JK.html)（2024-01-18）
- [Breaking Boundaries: AI CoScientist to Accelerate Science Research with Gabe Gomes, Professor at CMU](https://cogrev.qiaomu.ai/ep/095-breaking-boundaries-ai-coscientist-to-accelerate-science-res-_GbZn7.html)（2024-01-12）
- [AI-Powered Business Communications with Dan O'Connell, Chief AI & Strategy Officer at Dialpad](https://cogrev.qiaomu.ai/ep/094-ai-powered-business-communications-with-dan-o-connell-chief-c27107.html)（2024-01-09）
- [2028 AGI, New Architectures, and Aligning Superhuman Models with Shane Legg, Deepmind Founder, on The Dwarkesh Podcast](https://cogrev.qiaomu.ai/ep/093-2028-agi-new-architectures-and-aligning-superhuman-models-wi-fdf165.html)（2024-01-04）
- [Cutting-edge Neurotechnology with Nita Farahany in Conversation with Luisa Rodriguez on The 80,000 Hours Podcast](https://cogrev.qiaomu.ai/ep/092-cutting-edge-neurotechnology-with-nita-farahany-in-conversat-0a9d81.html)（2024-01-03）
- [AGI and Institutional Disruption: Sam Hammond in Conversation with FLI’s Gus Docker](https://cogrev.qiaomu.ai/ep/091-agi-and-institutional-disruption-sam-hammond-in-conversation-d1c9c9.html)（2023-12-30）
- [Nathan on The 80,000 Hours Podcast: AI Scouting, OpenAI's Safety Record, and Redteaming Frontier Models](https://cogrev.qiaomu.ai/ep/090-nathan-on-the-80-000-hours-podcast-ai-scouting-openai-s-safe-ZDiePf.html)（2023-12-27）
- [Emergency Pod: Mamba, Memory, and the SSM Moment](https://cogrev.qiaomu.ai/ep/089-emergency-pod-mamba-memory-and-the-ssm-moment-X5F2X4.html)（2023-12-22）
- [Using ChatGPT As a Copilot For Your Mind](https://cogrev.qiaomu.ai/ep/088-using-chatgpt-as-a-copilot-for-your-mind-qnzA9s.html)（2023-12-20）
- [AI Deception, Interpretability, and Affordances with Apollo Research CEO Marius Hobbhahn](https://cogrev.qiaomu.ai/ep/087-ai-deception-interpretability-and-affordances-with-apollo-re-wPayna.html)（2023-12-15）
- [Balaji Srinivasan on Polytheistic AI, Human-AI Symbiosis, and Prospects for AI Control](https://cogrev.qiaomu.ai/ep/086-balaji-srinivasan-on-polytheistic-ai-human-ai-symbiosis-and--9ROlC.html)（2023-12-12）
- [Hume CEO Alan Cowen on Creating Emotionally Aware AI](https://cogrev.qiaomu.ai/ep/085-hume-ceo-alan-cowen-on-creating-emotionally-aware-ai-3C1017.html)（2023-12-09）
- [What Biological & Social Systems Can Teach us About AI with Nora Ammann, Cofounder of PIBBSS Research](https://cogrev.qiaomu.ai/ep/084-what-biological-social-systems-can-teach-us-about-ai-with-no-mg53I4.html)（2023-12-07）
- [Imbue CTO Josh Albrecht on Creating AI Agents for Reasoning, Reliability, and Robustness](https://cogrev.qiaomu.ai/ep/083-imbue-cto-josh-albrecht-on-creating-ai-agents-for-reasoning-pQgL91.html)（2023-12-02）
- [Vercel CEO Guillermo Rauch on AI-Powered Coding, Software 2.0, and the AI Open Source Question](https://cogrev.qiaomu.ai/ep/082-vercel-ceo-guillermo-rauch-on-ai-powered-coding-software-2-0-KBGR5N.html)（2023-11-29）
- [Sam Altman Fired from OpenAI: NEW Insider Context on the Board’s Decision](https://cogrev.qiaomu.ai/ep/081-sam-altman-fired-from-openai-new-insider-context-on-the-boar-UdBMkj.html)（2023-11-22）
- [AI Discourse Deranged: Assessing LLM Generalization Takes and Polarizing Regulatory Debate](https://cogrev.qiaomu.ai/ep/080-ai-discourse-deranged-assessing-llm-generalization-takes-and-XJLykN.html)（2023-11-17）
- [Synthetic Data with Alex Watson, Founder of Gretel AI](https://cogrev.qiaomu.ai/ep/079-synthetic-data-with-alex-watson-founder-of-gretel-ai-uX8RgK.html)（2023-11-14）
- [OpenAI Invests $5M Into the Self-Driving Race with John Hayes, Cofounder of Ghost Autonomy](https://cogrev.qiaomu.ai/ep/078-openai-invests-5m-into-the-self-driving-race-with-john-hayes-2I3snM.html)（2023-11-09）
- [OpenAI DevDay: Beyond the Headlines with Logan Kilpatrick, OpenAI's Dev Relations Lead](https://cogrev.qiaomu.ai/ep/077-openai-devday-beyond-the-headlines-with-logan-kilpatrick-ope-WFM2pv.html)（2023-11-08）
- [Biden's Executive Order and AI Safety with Flo Crivello, Founder of Lindy AI](https://cogrev.qiaomu.ai/ep/076-biden-s-executive-order-and-ai-safety-with-flo-crivello-foun-e4EvKX.html)（2023-11-03）
- [Hardcore AI for History with Mark Humphries, Professor of History at Wilfred Laurier University](https://cogrev.qiaomu.ai/ep/075-hardcore-ai-for-history-with-mark-humphries-professor-of-his-IcIRbE.html)（2023-10-31）
- [AI Agents, VC Insights on AI, and Building in Public with Yohei Nakajima, Creator of BabyAGI](https://cogrev.qiaomu.ai/ep/074-ai-agents-vc-insights-on-ai-and-building-in-public-with-yohe-6DnnTD.html)（2023-10-26）
- [RAG & Beyond: Semantic Storage and Retrieval](https://cogrev.qiaomu.ai/ep/073-rag-beyond-semantic-storage-and-retrieval-fDmQnB.html)（2023-10-24）
- [The Future of the Transformer Part 2 with Trey Kollmer](https://cogrev.qiaomu.ai/ep/072-the-future-of-the-transformer-part-2-with-trey-kollmer-2QGQWW.html)（2023-10-20）
- [The Future of the Transformer Part 1 with Trey Kollmer | H100 Chips will Supercharge AI Hardware](https://cogrev.qiaomu.ai/ep/071-the-future-of-the-transformer-part-1-with-trey-kollmer-h100-SX_e9p.html)（2023-10-18）
- [The AI Scouting Report: Jailbreaks and Defense](https://cogrev.qiaomu.ai/ep/070-the-ai-scouting-report-jailbreaks-and-defense-VC_mGS.html)（2023-10-13）
- [AI is Supercharging Writers with Sudowrite's Founder, James Yu](https://cogrev.qiaomu.ai/ep/069-ai-is-supercharging-writers-with-sudowrite-s-founder-james-y-hDgGQy.html)（2023-10-10）
- [AI Engineers, Pendants, and Competition Between OpenAI and Developers with Swyx of Latent Space](https://cogrev.qiaomu.ai/ep/068-ai-engineers-pendants-and-competition-between-openai-and-dev-5EzAab.html)（2023-10-05）
- [Gunning for Google with Perplexity CEO Aravind Srinivas](https://cogrev.qiaomu.ai/ep/067-gunning-for-google-with-perplexity-ceo-aravind-srinivas-ix4_rd.html)（2023-10-03）
- [OpenAI, Amazon's Anthropic Investment, and the Roman Empire with Zvi Mowshowitz](https://cogrev.qiaomu.ai/ep/066-openai-amazon-s-anthropic-investment-and-the-roman-empire-wi-koGM30.html)（2023-09-29）
- [OpenAI Fine-Tuning Update, Acceleration Debate, and Bundling AI Services](https://cogrev.qiaomu.ai/ep/065-openai-fine-tuning-update-acceleration-debate-and-bundling-a-o6mOZ-.html)（2023-09-26）
- [Universal Jailbreaks with Zico Kolter, Andy Zou, and Asher Trockman](https://cogrev.qiaomu.ai/ep/064-universal-jailbreaks-with-zico-kolter-andy-zou-and-asher-tro-Bwltbh.html)（2023-09-22）
- [The Future of AI Security with Adam Wenchel, CEO of Arthur.ai](https://cogrev.qiaomu.ai/ep/063-the-future-of-ai-security-with-adam-wenchel-ceo-of-arthur-ai-o04b9s.html)（2023-09-19）
- [AI-Powered Filmmaking with Waymark's Stephen Parker and Josh Rubin](https://cogrev.qiaomu.ai/ep/062-ai-powered-filmmaking-with-waymark-s-stephen-parker-and-josh-c1pPiG.html)（2023-09-14）
- [Hollywood Strike Update and AI Roundup with Trey Kollmer](https://cogrev.qiaomu.ai/ep/061-hollywood-strike-update-and-ai-roundup-with-trey-kollmer-IoMyV_.html)（2023-09-12）
- [Why a16z Built a Town for AI People](https://cogrev.qiaomu.ai/ep/060-why-a16z-built-a-town-for-ai-people-uxbfIg.html)（2023-09-07）
- [Cryptography Will Revolutionize AI Data Privacy with Daniel Kang](https://cogrev.qiaomu.ai/ep/059-cryptography-will-revolutionize-ai-data-privacy-with-daniel-O9sWKx.html)（2023-09-05）
- [How Hugging Face raised $235M](https://cogrev.qiaomu.ai/ep/058-how-hugging-face-raised-235m-XWdV5P.html)（2023-08-31）
- [E57: Google's PaLM 2 with Paige Bailey](https://cogrev.qiaomu.ai/ep/057-e57-google-s-palm-2-with-paige-bailey-K-XYxL.html)（2023-08-29）
- [E56: OpenAI’s GPT3.5 Turbo and the State of AI in Coding, Healthcare, and Education](https://cogrev.qiaomu.ai/ep/056-e56-openais-gpt3-5-turbo-and-the-state-of-ai-in-coding-healt-O3capV.html)（2023-08-25）
- [E55: Google’s Multimodal Med-PaLM with Vivek Natarajan and Tao Tu](https://cogrev.qiaomu.ai/ep/055-e55-googles-multimodal-med-palm-with-vivek-natarajan-and-tao-2RMpqh.html)（2023-08-23）
- [E54: The Presentation Revolution with Jon Noronha, Co-Founder of Gamma](https://cogrev.qiaomu.ai/ep/054-e54-the-presentation-revolution-with-jon-noronha-co-founder-d7dde4.html)（2023-08-17）
- [E53: The AI Revolution in Education with Shawn Jansepar, Director of Engineering at Khan Academy](https://cogrev.qiaomu.ai/ep/053-e53-the-ai-revolution-in-education-with-shawn-jansepar-direc-ati_AC.html)（2023-08-15）
- [E52: Advancements in AI: Updating the Scouting Report, Task Automation, and Google Breakthroughs](https://cogrev.qiaomu.ai/ep/052-e52-advancements-in-ai-updating-the-scouting-report-task-aut-2GuURc.html)（2023-08-10）
- [E51: Replit's VP of AI, Michele Catasta, on Artificial Developer Intelligence](https://cogrev.qiaomu.ai/ep/051-e51-replit-s-vp-of-ai-michele-catasta-on-artificial-develope-u4l6Gg.html)（2023-08-08）
- [E50: The Future of Software Development with Tyler Angert of Replit](https://cogrev.qiaomu.ai/ep/050-e50-the-future-of-software-development-with-tyler-angert-of-uEMN51.html)（2023-08-04）
- [E49: Open AI, Anthropic, and Meta | Analyzing the AI Frontier with Zvi](https://cogrev.qiaomu.ai/ep/049-e49-open-ai-anthropic-and-meta-analyzing-the-ai-frontier-wit-SM4q-Q.html)（2023-08-02）
- [E48: Mechanizing Mechanistic Interpretability with Arthur Conmy](https://cogrev.qiaomu.ai/ep/048-e48-mechanizing-mechanistic-interpretability-with-arthur-con-Y5Pzch.html)（2023-07-27）
- [E47: GPUMaxing with Dr. Ronen Dar of Run:ai](https://cogrev.qiaomu.ai/ep/047-e47-gpumaxing-with-dr-ronen-dar-of-run-ai-RvD1R7.html)（2023-07-25）
- [E46: AI Scouting Report - Part 2 of 3: Implementation Trends [Bonus YouTube Episode]](https://cogrev.qiaomu.ai/ep/046-e46-ai-scouting-report-part-2-of-3-implementation-trends-bon-ovm4Mb.html)（2023-07-19）
- [E45: The AI Copilot Revolution with Div Garg of MULTI·ON](https://cogrev.qiaomu.ai/ep/045-e45-the-ai-copilot-revolution-with-div-garg-of-multion-PR2Mdl.html)（2023-07-13）
- [E44: The AI Safety Debates with Zvi Mowshowitz](https://cogrev.qiaomu.ai/ep/044-e44-the-ai-safety-debates-with-zvi-mowshowitz-5yM7fI.html)（2023-07-11）
- [E43: AI Scouting Report – Part 1 of 3: Fundamentals [Bonus YouTube Episode]](https://cogrev.qiaomu.ai/ep/043-e43-ai-scouting-report-part-1-of-3-fundamentals-bonus-youtub-0hvtiV.html)（2023-07-06）
- [E42: The Autonomy Revolution with Hayk Martiros of Skydio](https://cogrev.qiaomu.ai/ep/042-e42-the-autonomy-revolution-with-hayk-martiros-of-skydio-aaLIxW.html)（2023-07-04）
- [E41: [Bonus Episode - Latent.space] Building the AI x UX Scenius with Linus Lee of Notion AI](https://cogrev.qiaomu.ai/ep/041-e41-bonus-episode-latent-space-building-the-ai-x-ux-scenius-001272.html)（2023-07-01）
- [E40: Meta's MEGABYTE Revolution with Lili Yu of Meta AI](https://cogrev.qiaomu.ai/ep/040-e40-meta-s-megabyte-revolution-with-lili-yu-of-meta-ai-8EIqHF.html)（2023-06-29）
- [E39: Seeing is Believing with MIT’s Ziming Liu](https://cogrev.qiaomu.ai/ep/039-e39-seeing-is-believing-with-mits-ziming-liu-VKxwlx.html)（2023-06-27）
- [E38: The Virtual Biopsy Revolution with Dr. Tanishq Mathew Abraham (Part 2 of 2)](https://cogrev.qiaomu.ai/ep/038-e38-the-virtual-biopsy-revolution-with-dr-tanishq-mathew-abr-Vpt9z3.html)（2023-06-22）
- [E37: The Mind-Reading Revolution with Dr. Tanishq Mathew Abraham (Part 1 of 2)](https://cogrev.qiaomu.ai/ep/037-e37-the-mind-reading-revolution-with-dr-tanishq-mathew-abrah-JY7K-h.html)（2023-06-20）
- [E36: Your Model, Your Weights with MosaicML's Abhi Venigalla and Jonathan Frankle](https://cogrev.qiaomu.ai/ep/036-e36-your-model-your-weights-with-mosaicml-s-abhi-venigalla-a-5_qAuS.html)（2023-06-16）
- [E35: Stability AI’s Emad Mostaque and Slow Ventures’ Sam Lessin Discuss Investing in AI](https://cogrev.qiaomu.ai/ep/035-e35-stability-ais-emad-mostaque-and-slow-ventures-sam-lessin-mOOYJO.html)（2023-06-13）
- [E34: The Consumer Rights Revolution with Joshua Browder of DoNotPay](https://cogrev.qiaomu.ai/ep/034-e34-the-consumer-rights-revolution-with-joshua-browder-of-do-ejB2kv.html)（2023-06-09）
- [E33: The Tiny Model Revolution with Ronen Eldan and Yuanzhi Li of Microsoft Research](https://cogrev.qiaomu.ai/ep/033-e33-the-tiny-model-revolution-with-ronen-eldan-and-yuanzhi-l-mv3SIg.html)（2023-06-06）
- [E32: [Bonus Episode - The AI Breakdown] Can OpenAI's New GPT Training Model Solve Math and AI Alignment At the Same Time?](https://cogrev.qiaomu.ai/ep/032-e32-bonus-episode-the-ai-breakdown-can-openai-s-new-gpt-trai-aaeb86.html)（2023-06-03）
- [E31: Where are the Moats in AI? With Nathan Labenz and Erik Torenberg](https://cogrev.qiaomu.ai/ep/031-e31-where-are-the-moats-in-ai-with-nathan-labenz-and-erik-to-yVezX3.html)（2023-06-01）
- [E30: The AI Revolution in Hollywood? With WGA writers Trey Kollmer, Sophia Lear, and Garrett Schabb](https://cogrev.qiaomu.ai/ep/030-e30-the-ai-revolution-in-hollywood-with-wga-writers-trey-kol-BkXiQi.html)（2023-05-30）
- [E29: The AI Chip Revolution with Andrew Feldman of Cerebras](https://cogrev.qiaomu.ai/ep/029-e29-the-ai-chip-revolution-with-andrew-feldman-of-cerebras-JjQkNz.html)（2023-05-25）
- [E28: Keeping the AI Revolution on the Rails with Shreya Rajpal of Guardrails AI](https://cogrev.qiaomu.ai/ep/028-e28-keeping-the-ai-revolution-on-the-rails-with-shreya-rajpa-97plVw.html)（2023-05-23）
- [E27: “Google’s Med-PaLM and Med-PaLM2 with Vivek Natarajan”](https://cogrev.qiaomu.ai/ep/027-e27-googles-med-palm-and-med-palm2-with-vivek-natarajan-nPBd7i.html)（2023-05-21）
- [E26: [Bonus Episode] Connor Leahy on AGI, GPT-4, and Cognitive Emulation w/ FLI Podcast](https://cogrev.qiaomu.ai/ep/026-e26-bonus-episode-connor-leahy-on-agi-gpt-4-and-cognitive-em-7BqcdY.html)（2023-05-19）
- [E25: Revolutionizing Patient Care with Neal Khosla of Curai Health](https://cogrev.qiaomu.ai/ep/025-e25-revolutionizing-patient-care-with-neal-khosla-of-curai-h-Sqpkt8.html)（2023-05-16）
- [E24: The AI Revolution in Medicine with Dr. Isaac Kohane of Harvard Medical School](https://cogrev.qiaomu.ai/ep/024-e24-the-ai-revolution-in-medicine-with-dr-isaac-kohane-of-ha-pS5Vye.html)（2023-05-11）
- [E23: Scouting the AI Revolution with Robert Scoble and "Ben's Bites" creator Ben Tossell](https://cogrev.qiaomu.ai/ep/023-e23-scouting-the-ai-revolution-with-robert-scoble-and-ben-s-7pAq5I.html)（2023-05-09）
- [E22: Helping Businesses Use AI with Rachel Woods of The AI Exchange](https://cogrev.qiaomu.ai/ep/022-e22-helping-businesses-use-ai-with-rachel-woods-of-the-ai-ex--U6Wrg.html)（2023-05-04）
- [E21: VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors Podcast](https://cogrev.qiaomu.ai/ep/021-e21-vc-insights-on-investing-in-artificial-intelligence-with-2tMDBa.html)（2023-05-02）
- [E20: The Great Implementation with Raza Habib, CEO of Humanloop](https://cogrev.qiaomu.ai/ep/020-e20-the-great-implementation-with-raza-habib-ceo-of-humanloo-EKKJrR.html)（2023-04-27）
- [E19: The AI Agent Revolution with Matt Welsh of Fixie.ai](https://cogrev.qiaomu.ai/ep/019-e19-the-ai-agent-revolution-with-matt-welsh-of-fixie-ai-MHmd8h.html)（2023-04-25）
- [E18: Why Jailbreaking ChatGPT Is A Public Good with Alex Albert of The Prompt Report](https://cogrev.qiaomu.ai/ep/018-e18-why-jailbreaking-chatgpt-is-a-public-good-with-alex-albe-2WwL6V.html)（2023-04-20）
- [E17: The Prompt Engineering Revolution with Riley Goodside](https://cogrev.qiaomu.ai/ep/017-e17-the-prompt-engineering-revolution-with-riley-goodside-zg3H-9.html)（2023-04-18）
- [E16: Pausing the AI Revolution? With Technologist Jaan Tallinn](https://cogrev.qiaomu.ai/ep/016-e16-pausing-the-ai-revolution-with-technologist-jaan-tallinn-R78mbt.html)（2023-04-13）
- [E15: Will GPT-4 Cause Economic Transformation?](https://cogrev.qiaomu.ai/ep/015-e15-will-gpt-4-cause-economic-transformation-HaBLuM.html)（2023-04-11）
- [E14: The Reasoning Revolution with Ought's Jungwon Byun and Andreas Stuhlmüller](https://cogrev.qiaomu.ai/ep/014-e14-the-reasoning-revolution-with-ought-s-jungwon-byun-and-a-eZF7tQ.html)（2023-04-06）
- [E13: AI open letter debate, "pausers", and what we're all scared of with AI with Anton Troynikov and Flo Crivello [Moment of Zen]](https://cogrev.qiaomu.ai/ep/013-e13-ai-open-letter-debate-pausers-and-what-we-re-all-scared-p6T1QC.html)（2023-04-04）
- [E12: The Robot Revolution with Keerthana Gopalakrishnan of Google Robotics](https://cogrev.qiaomu.ai/ep/012-e12-the-robot-revolution-with-keerthana-gopalakrishnan-of-go-5tlQhg.html)（2023-03-30）
- [E11: OpenAI's GPT-4 Discussion with Nathan Labenz and Erik Torenberg](https://cogrev.qiaomu.ai/ep/011-e11-openai-s-gpt-4-discussion-with-nathan-labenz-and-erik-to-oLiheM.html)（2023-03-28）
- [E10: The AI Voice Revolution with Mahmoud Felfel of Play.ht](https://cogrev.qiaomu.ai/ep/010-e10-the-ai-voice-revolution-with-mahmoud-felfel-of-play-ht-9YtxBn.html)（2023-03-24）
- [E9: The AI Assistant Revolution with Flo Crivello of Lindy.AI](https://cogrev.qiaomu.ai/ep/009-e9-the-ai-assistant-revolution-with-flo-crivello-of-lindy-ai-tGKULK.html)（2023-03-22）
- [E8: GPT4 - AI Unleashed on ChinaTalk Podcast](https://cogrev.qiaomu.ai/ep/008-e8-gpt4-ai-unleashed-on-chinatalk-podcast-PlbdcP.html)（2023-03-17）
- [E7: In Search of Truth with Aravind Srinivas of Perplexity AI](https://cogrev.qiaomu.ai/ep/007-e7-in-search-of-truth-with-aravind-srinivas-of-perplexity-ai-YSYQZZ.html)（2023-03-14）
- [E6: The Computer Vision Revolution with Junnan Li and Dongxu Li of BLIP and BLIP2](https://cogrev.qiaomu.ai/ep/006-e6-the-computer-vision-revolution-with-junnan-li-and-dongxu-zTr5vD.html)（2023-03-09）
- [E5: The Embedding Revolution: Anton Troynikov on Chroma, Stable Attribution, and future of AI](https://cogrev.qiaomu.ai/ep/005-e5-the-embedding-revolution-anton-troynikov-on-chroma-stable-ogy37C.html)（2023-03-02）
- [E4: The AI Moment with Amjad Masad, Flo Crivello, Dan Romero, and Antonio Garcia Martinez](https://cogrev.qiaomu.ai/ep/004-e4-the-ai-moment-with-amjad-masad-flo-crivello-dan-romero-an-oQaA3W.html)（2023-02-23）
- [E3: The Empathy Revolution with Eugenia Kuyda of Replika](https://cogrev.qiaomu.ai/ep/003-e3-the-empathy-revolution-with-eugenia-kuyda-of-replika-SFKA7T.html)（2023-02-17）
- [E2: Why AI Will Cause a Cognitive Revolution with Nathan Labenz and Erik Torenberg](https://cogrev.qiaomu.ai/ep/002-e2-why-ai-will-cause-a-cognitive-revolution-with-nathan-labe-iGD2TC.html)（2023-02-10）
- [E1: The Pixel Revolution with Playground AI 's Suhail Doshi](https://cogrev.qiaomu.ai/ep/001-e1-the-pixel-revolution-with-playground-ai-s-suhail-doshi-Waii0i.html)（2023-02-03）